deletions | additions       diff --git a/VR-blindsightNotes/blindsight.bib b/VR-blindsightNotes/blindsight.bib  new file mode 100644  index 0000000..144d699  --- /dev/null  +++ b/VR-blindsightNotes/blindsight.bib 
...
@Article{ajina.EA:15motion,  author = {Sara Ajina and Christopher Kennard and Geraint Rees  and Holly Bridge},  title = {Motion area {V5/MT+} response to global motion in  the absence of {V1} resembles early visual cortex},  journal = {Brain},  year = 2015,  volume = 138,  pages = {164--178},  note = {dx.doi.org/10.1093/brain/awu328},  abstract = {Motion area V5/MT+ shows a variety of characteristic  visual responses, often linked to perception, which  are heavily influenced by its rich connectivity with  the primary visual cortex (V1). This human motion  area also receives a number of inputs from other  visual regions, including direct subcortical  connections and callosal connections with the  contralateral hemisphere. Little is currently known  about such alternative inputs to V5/MT+ and how they  may drive and influence its activity. Using  functional magnetic resonance imaging, the response  of human V5/MT+ to increasing the proportion of  coherent motion was measured in seven patients with  unilateral V1 damage acquired during adulthood, and  a group of healthy age-matched controls. When V1 was  damaged, the typical V5/MT+ response to increasing  coherence was lost. Rather, V5/MT+ in patients  showed a negative trend with coherence that was  similar to coherence-related activity in V1 of  healthy control subjects. This shift to a  response-pattern more typical of early visual cortex  suggests that in the absence of V1, V5/MT+ activity  may be shaped by similar direct subcortical  input. This is likely to reflect intact residual  pathways rather than a change in connectivity, and  has important implications for blindsight  function. It also confirms predictions that V1 is  critically involved in normal V5/MT+ global motion  processing, consistent with a convergent model of V1  input to V5/MT+. Historically, most attempts to  model cortical visual responses do not consider the  contribution of direct subcortical inputs that may  bypass striate cortex, such as input to V5/MT+. We  have shown that the signal change driven by these  non-striate pathways can be measured, and suggest  that models of the intact visual system may benefit  from considering their contribution.}  }  @Article{ajina.EA:15contr,  author = {Sara Ajina and Geraint Rees and Christopher Kennard  and Holly Bridge},  title = {Abnormal Contrast Responses in the Extrastriate  Cortex of Blindsight Patients},  journal = JoN,  year = 2015,  volume = 35,  number = 21,  pages = {8201--8213},  abstract = {When the human primary visual cortex (V1) is  damaged, the dominant geniculo-striate pathway can  no longer convey visual information to the occipital  cortex. However, many patients with such damage  retain some residual visual function that must rely  on an alternative pathway directly to extrastriate  occipital regions. This residual vision is most  robust for moving stimuli, suggesting a role for  motion area hMT . However, residual vision also  requires high-contrast stimuli, which is  inconsistent with hMT sensitivity to contrast in  which even low-contrast levels elicit near-maximal  neural activation. We sought to investigate this  discrepancy by measuring behavioral and neural  responses to increasing contrast in patients with V1  damage. Eight patients underwent behavioral testing  and functional magnetic resonance imaging to record  contrast sensitivity in hMT of their damaged  hemisphere, using Gabor stimuli with a spatial  frequency of 1 cycle/°. The responses from hMT of  the blind hemisphere were compared with hMT and V1  responses in the sighted hemisphere of patients and  a group of age-matched controls. Unlike hMT , neural  responses in V1 tend to increase linearly with  increasing contrast, likely reflecting a dominant  parvocellular channel input. Across all patients,  the responses in hMT of the blind hemisphere no  longer showed early saturation but increased  linearly with contrast. Given the spatiotemporal  parameters used in this study and the known direct  subcortical projections from the koniocellular  layers of the lateral geniculate nucleus to hMT , we  propose that this altered contrast sensitivity in  hMT could be consistent with input from the  koniocellular pathway.}  }  @Article{ajina.EA:15dMRI,  author = {Sara Ajina and Franco Pestilli and Ariel Rokem and  Christopher Kennard and Holly Bridge},  title = {Human blindsight is mediated by an intact  geniculo-extrastriate pathway},  journal = {eLIFE},  year = 2015,  volume = 4,  pages = {e08935},  note = {10.7554/eLife.08935},  abstract = {Although damage to the primary visual cortex (V1)  causes hemianopia, many patients retain some  residual vision; known as blindsight. We show that  blindsight may be facilitated by an intact  white-matter pathway between the lateral geniculate  nucleus and motion area hMT+. Visual psychophysics,  diffusion-weighted magnetic resonance imaging and  fibre tractography were applied in 17 patients with  V1 damage acquired during adulthood and 9  age-matched controls. Individuals with V1 damage  were subdivided into blindsight positive (preserved  residual vision) and negative (no residual vision)  according to psychophysical performance. All  blindsight positive individuals showed intact  geniculo-hMT+ pathways, while this pathway was  significantly impaired or not measurable in  blindsight negative individuals. Two white matter  pathways previously implicated in blindsight: (i)  superior colliculus to hMT+ and (ii) between hMT+ in  each hemisphere were not consistently present in  blindsight positive cases. Understanding the visual  pathways crucial for residual vision may direct  future rehabilitation strategies for hemianopia  patients. }  }  @Article{anwar.EA:09,  author = {Muhammad Nabeel Anwar and Laura Bonzano and Davide  Rossi Sebastiano and Luca Roccatagliata and Giovanni  Gualniera and Paolo Vitali and Carla Ogliastro and  Luciano Spadavecchia and Guido Rodriguez and  Vittorio Sanguineti and Pietro Morasso and Fabio  Bandini},  title = { Real-time artifact filtering in continuous  {VEPs/fMRI} recording},  journal = JNMeth,  year = 2009,  volume = 184,  pages = 184,  abstract = {Continuous recording of Visual Evoked Potentials  (VEPs) and functional Magnetic Resonance Imaging  (fMRI) exploits the VEPs high temporal resolution  and the fMRI high spatial resolution. In this work,  we present a new method of continuous VEPs/fMRI  recording to study visual function in seven normal  subjects. Our real-time artifact filtering is  characterized by a procedure based on an analytical  study of echo-planar imaging (EPI) sequence  parameters related electro-encephalogram  (EEG)-artifact shapes. The magnetic field artifacts  were minimized by using a dedicated amagnetic device  and by a subtraction algorithm that takes into  account the EPI sequence parameters. No significant  decrease in signal-to-noise ratio was observed in  case of EEG recording simultaneously with MR  acquisition; similarly, transient and steady-state  VEPs parameters were comparable during fMRI  acquisition and in the off-phase of fMRI  recording. We also applied this method to one  patient with optic neuritis, and, compared with  controls, found different results. We suggest that  our technique can be reliably used to investigate  the function of human visual cortex and properly  correlate the electrophysiological and functional  neuroimaging related changes.}  }  @Article{azzopardi.hock:11,  author = {Paul Azzopardi and Howard S. Hock},  title = {Illusory motion perception in blindsight},  journal = PNAS,  year = 2011,  volume = 108,  number = 2,  pages = {876--881},  abstract = {Motion detection is typically spared in blindsight,  which results from damage to the striate cortex  (area V1) of the brain that is sufficient to  eliminate conscious visual awareness and severely  reduce sensitivity to luminance contrast, especially  for high spatial and low temporal frequencies. Here  we show that the discrimina- tion of motion  direction within cortically blind fields is not  attributable to feature tracking (the detection of  changes in position or shape), but is due instead to  the detection of first- order motion energy  (spatiotemporal changes in luminance). The key to  this finding was a version of the line motion  illusion entailing reverse-phi motion in which  opposing motion directions are simultaneously cued  by motion energy and changes in stimulus shape. In  forced-choice tests, a blindsighted test subject  selected the direction cued by shape change when the  stimulus was presented in his intact field, but  reliably selected the direction cued by motion  energy when the same stimulus was presented in his  blind field, where relevant position information was  either inaccessible or invalid. Motion energy has  been characterized as objectless, so reliance on  motion energy detection is consistent with impaired  access to shape information in blindsight. The  dissociation of motion direction by visual field  (cortically blind vs. intact) provides evidence that  two pathways from the retina to MT/V5 (the cortical  area specialized for motion perception) are  functionally distinct: the retinogeniculate pathway  through V1 is specialized for feature-based motion  perception, whereas the retinocollicular pathway,  which bypasses V1, is specialized for detecting  motion energy.}  }  @Article{he.EA:11,  author = {Bin He and Yakang Dai and Laura Astolfi and Fabio  Babiloni and Han Yuan and Lin Yang},  title = {{eConnectome: A MATLAB} toolbox for mapping and  imaging of brain functional connectivity},  journal = JNMeth,  year = 2011,  volume = 195,  number = 2,  pages = {261--269},  abstract = {We have developed a MATLAB-based toolbox,  eConnectome (electrophysiological connectome), for  mapping and imaging functional connectivity at both  the scalp and cortical levels from the  electroencephalogram (EEG), as well as from the  electrocorticogram (ECoG). Graphical user interfaces  were designed for interactive and intuitive use of  the toolbox. Major functions of eConnectome include  EEG/ECoG preprocessing, scalp spatial mapping,  cortical source estimation, connectivity analysis,  and visualization. Granger causality measures such  as directed transfer function and adaptive directed  transfer function were implemented to estimate the  directional interactions of brain functional  networks, over the scalp and cortical sensor  spaces. Cortical current density inverse imaging was  implemented using a generic realistic geometry  brain–head model from scalp EEGs. Granger causality  could be further estimated over the cortical source  domain from the inversely reconstructed cortical  source signals as derived from the scalp EEG. Users  may implement other connectivity estimators in the  framework of eConnectome for various  applications. The toolbox package is open-source and  freely available at http://econnectome.umn.edu under  the GNU general public license for noncommercial and  academic uses.}  }  @Article{hindiattar.EA:10,  author = {Catherine {Hindi Attar} and S{\/o}ren K. Andersen and  Matthias M. M{\"u}ller},  title = {Time course of affective bias in visual attention:  Convergent evidence from steady-state visual evoked  potentials and behavioral data},  journal = NI,  year = 2010,  volume = 53,  pages = {1326--1333},  abstract = {Selective attention to a primary task can be biased  by the occurrence of emotional distractors that  involuntary attract attention due to their intrinsic  stimulus significance. What is largely unknown is  the time course and magnitude of competitive  interactions between a to-be-attended foreground  task and emotional distractors. We used pleasant,  unpleasant and neutral pictures from the  International Affective Picture System (IAPS) that  were either presented in intact or phase-scrambled  form. Pictures were superimposed by a flickering  display of moving random dots, which constituted the  primary task and enabled us to record steady-state  visual evoked potentials (SSVEPs) as a continuous  measure of attentional resource allocation directed  to the task. Subjects were required to attend to the  dots and to detect short intervals of coherent  motion while ignoring the background pictures. We  found that pleasant and unpleasant relative to  neutral pictures more strongly influenced  task-related processing as reflected in a  significant decrease in SSVEP amplitudes and target  detection rates, both covering a time window of  several hundred milliseconds. Strikingly, the effect  of semantic relative to phase-scrambled pictures on  task-related activity was much larger, emerged  earlier and lasted longer in time compared to the  specific effect of emotion. The observed differences  in size and duration of time courses of semantic and  emotional picture processing strengthen the  assumption of separate functional mechanisms for  both processes rather than a general boosting of  neural activity in favor of emotional stimulus  processing.}  }  @Article{bagattini.EA:15,  author = {Chiara Bagattini and Chiara Mazzi and Silvia  Savazzi},  title = {Waves of awareness for occipital and parietal  phosphenes perception},  journal = {Neuropsychologia},  year = 2015,  volume = 70,  pages = {114--125},  abstract = {Transcranial magnetic stimulation (TMS) of the  occipital cortex is known to induce visual  sensations, i.e. phosphenes, which appear as fl  ashes of light in the absence of an external  stimulus. Recent studies have shown that TMS can  produce phosphenes also when the intraparietal  sulcus (IPS) is stimulated. The main question  addressed in this paper is whether parietal  phosphenes are generated directly by local me-  chanisms or emerge through indirect activation of  other visual areas. Electroencephalographic (EEG)  signals were recorded while stimulating left  occipital or parietal cortices inducing phosphene  perception in healthy participants and in a  hemianopic patient who suffered from complete  destruction of the early visual cortex of the left  hemisphere. Results in healthy participants showed  that the onset of phosphene perception induced by  occipital TMS correlated with differential cortical  activity in temporal sites while the onset of  phosphene perception induced by parietal TMS  correlated with differential cortical activity in  the stimulated parietal site. Moreover, IPS-TMS of  the lesioned hemisphere of the hemianopic patient  with a complete lesion to V1 showed again that the  onset of phosphene perception correlated with  differential cortical activity in the stimulated  parietal site. The present data seem thus to suggest  that temporal and parietal cortices can serve as  different local early gatekeepers of perceptual  awareness and that activity in the occipital cortex,  although being relevant for perception in general,  is not part of the neural bases of the perceptual  awareness of phosphenes.}  }  @Article{bandettini.cox:00,  author = {Peter A. Bandettini and Robert W. Cox},  title = {Event-Related {fMRI} Contrast When Using Constant  Interstimulus Interval: Theory and Experiment},  journal = MRM,  year = 2000,  volume = 43,  pages = {540--548},  annote = {Event-related functional magnetic resonance imaging  (ER- fMRI) involves the mapping of averaged  hemodynamic changes resulting from repeated, brief (  < 3 sec) brain activation epi- sodes. In this paper,  two issues regarding constant-interstimu- lus  interval ER-fMRI were addressed. First, the optimal  inter- stimulus interval (ISI), given a stimulus  duration (SD), was de- termined. Second, the  statistical power of ER-fMRI relative to that of a  blocked-design paradigm was determined. Experimen-  tally, it was found that with a 2-sec SD, the  optimal ISI is 12 to 14 sec. Theoretically, the  optimal repetition interval ( T opt 5 ISI 1 SD) is  12 to 14 sec for stimuli of 2 sec or less. For  longer stimuli, T opt is 8+2 * SD. At the optimal  ISI for SD=2 sec, the experimentally determined  functional contrast of ER-fMRI was only 2 35% lower  than that of blocked-design fMRI. Simulations that  assumed a linear system demonstrated an  event-related functional contrast that was 2 65%  lower than that of the blocked design. These  differences between simulated and ex- perimental  contrast suggest that the ER-fMRI amplitude is  greater than that predicted by a linear  shift-invariant system.}  }  @Article{barbur.EA:93,  author = {J. L. Barbur and J. D. G. Watson and R. S. J.  Frackowiak and S. Zeki},  title = {Conscious visual perception without {VI}},  journal = {Brain},  year = 1993,  volume = 116,  pages = {1293--1302},  abstract = {We used the technique of PET to determine whether  visual signals reach visual area V5, specialized for  visual motion, when a human patient, blinded by a  lesion in area VI, discriminates the direction of  motion of visual stimuli and shows, through his  verbal reports, that he is consciously aware of both  the nature of the visual stimulus and its direction  of motion. The results showed that area V5 was  active without a parallel activation of area VI,  implying that the visual input can reach V5 without  passing first through VI and that such an input is  sufficient for both the discrimination and the  conscious awareness of the visual stimulus.}  }  @Article{bayram.EA:11,  author = {Ali Bayram and Zubeyir Bayraktaroglu and Esin  Karahan and Basri Erdogan and Basar Bilgic and Muge  Özker and Itir Kasikci and Adil D. Duru and Ahmet  Ademoglu and Cengizhan {\"O}zt{\"u}rk and Kemal  Arikan and Nevzat Tarhan and Tamer Demiralp},  title = {Simultaneous {EEG/fMRI} Analysis of the Resonance  Phenomena in {Steady-State Visual Evoked Responses}},  journal = CEEGN,  year = 2011,  volume = 42,  number = 2,  pages = {98--106},  annote = {The stability of the steady-state visual evoked  potentials (SSVEPs) across trials and subjects makes  them a suitable tool for the investigation of the  visual system. The reproducible pattern of the  frequency charac- teristics of SSVEPs shows a global  amplitude maximum around 10 Hz and additional local  maxima around 20 and 40 Hz, which have been argued  to represent resonant behavior of damped neuronal  oscillators. Simultaneous  electroencephalogram/functional magnetic reson- ance  imaging (EEG/fMRI) measurement allows testing of the  reson- ance hypothesis about the frequency-selective  increases in SSVEP amplitudes in human subjects,  because the total synaptic activity that is  represented in the fMRI-Blood Oxygen Level Dependent  (fMRI- BOLD) response would not increase but get  synchronized at the resonance frequency. For this  purpose, 40 healthy volunteers were visually  stimulated with flickering light at systematically  varying frequencies between 6 and 46 Hz, and the  correlations between SSVEP amplitudes and the BOLD  responses were computed. The SSVEP frequency  characteristics of all subjects showed 3 frequency  ranges with an amplitude maximum in each of them,  which roughly correspond to alpha, beta and gamma  bands of the EEG. The correlation maps between BOLD  responses and SSVEP amplitude changes across the  different stimulation frequencies within each  frequency band showed no significant correlation in  the alpha range, while significant correlations were  obtained in the primary visual area for the beta and  gamma bands. This non-linear relationship between  the surface recorded SSVEP amplitudes and the BOLD  responses of the visual cortex at stimulation  frequencies around the alpha band supports the view  that a resonance at the tuning frequency of the  thalamo-cortical alpha oscillator in the visual  system is responsible for the global amplitude  maximum of the SSVEP around 10 Hz. Information  gained from the SSVEP/fMRI analyses in the present  study might be extrapolated to the EEG/fMRI analysis  of the transient event-related potentials (ERPs) in  terms of expecting more reliable and consistent  correlations between EEG and fMRI responses, when  the analyses are carried out on evoked or induced  oscillations (spectral perturbations) in separate  frequency bands instead of the time-domain ERP  peaks.}  }  @Article{becker.EA:10,  author = {Stefanie I. Becker and Charles L. Folk and Roger  W. Remington},  title = {The Role of Relational Information in Contingent  Capture},  journal = JEPhpp,  year = 2010,  volume = 36,  number = 6,  pages = {1460--1476},  abstract = {On the contingent capture account, top-down  attentional control settings restrict involuntary  attentional capture to items that match the features  of the search target. Attention capture is  involuntary, but contingent on goals and  intentions. The observation that only target-similar  items can capture attention has usually been taken  to show that the content of the attentional control  settings consists of specific feature values. In  contrast, the present study demonstrates that the  top-down target template can include information  about the relationship between the target and  nontarget features (e.g., redder, darker,  larger). Several spatial cuing experiments show that  a singleton cue that is less similar to the target  but that shares the same relational property that  distinguishes targets from nontargets can capture  attention to the same extent as cues that are  similar to the target. Moreover, less similar cues  can even capture attention more than cues that are  identical to the target when they are relationally  better than identical cues. The implications for  current theories of attentional capture and  attentional guidance are discussed.}  }  @Article{benoni.tsal:10,  author = {Hanna Benoni and Yehoshua Tsal},  title = {Where have we gone wrong? Perceptual load does not  affect selective attention},  journal = VisRes,  year = 2010,  volume = 50,  pages = {1292--1298},  abstract = {The theory of perceptual load (Lavie & Tsal, 1994)  proposes that with low load in relevant processing  left over resources spill over to process irrelevant  distractors. Interference could only be prevented  under High-Load Conditions where relevant processing  exhausts attentional resources. The theory is based  primarily on the finding that distractor  interference obtained in low load displays, when the  target appears alone, is eliminated in high load  displays when it is embedded among neutral  letters. However, a possible alternative  interpretation of this effect is that the distractor  is similarly processed in both displays, yet its  interference in the large displays is diluted by the  presence of the neutral letters. We separated the  possible effects of load and dilution by adding  dilution displays that were high in dilution and low  in perceptual load. In the first experiment these  displays contained as many letters as the high load  displays, but their neutral letters were clearly  distinguished from the target, thereby allowing for  a low load processing mode. In the second experiment  we presented identical multicolor displays in the  Dilution and High-Load Conditions. However, in the  former the target color was known in advance  (thereby preserving a low load processing mode)  whereas in the latter it was not. In both  experiments distractor interference was completely  eliminated under the Dilution Condition. Thus, it is  dilution not perceptual load affecting distractor  processing.}  }  @Article{benson.EA:98,  author = {Philip J. Benson and Kun Guo and Colin Blakemore},  title = {Direction discrimination of moving gratings and  plaids and coherence in dot displays without primary  visual cortex ({V1}).},  journal = EJN,  year = 1998,  volume = 10,  number = 12,  pages = {3767--3772},  abstract = {We present new experimental observations of G.Y., a  well-tested patient with unilateral loss of primary  visual cortex. We stimulated G.Y.'s blind hemifield  using first- and second-order motion stimuli at  velocities around psychophysical threshold. Using a  dual response paradigm (awareness level of visual  motion, motion direction discrimination)  psychophysical performance improved with increasing  velocity and dot coherence. We were also able to  influence directly G.Y.'s performance for the better  and at will, by placing the emphasis solely on  direction discrimination. In the absence of V1,  graduated detection and discrimination of stimuli  known to activate both V1 and extrastriate motion  areas MT/V5 and MST is still possible. These results  are in line with residual visual processing but did  not show evidence of unconscious processing of  motion stimuli characteristic of 'blindsight'.}  }  @Article{bertoldi.EA:15,  author = {F. de Bertoldi and L. Finos and M. Maieron and  L. Weis and M. Campanella and T. Ius and L. Fadiga},  title = {Improving the reliability of single-subject {fMRI}  by weighting intra-run variability},  journal = NI,  year = 2015,  volume = 114,  pages = {287--293},  abstract = {At present, functional magnetic resonance imaging  (fMRI) is one of the most useful methods of studying  cogni- tive processes in the human brain in vivo,  both for basic science and clinical goals. Although  neuroscience studies often rely on group analysis,  clinical applications must investigate single  subjects (patients) only. Particularly for the  latter, issues regarding the reliability of fMRI  readings remain to be resolved. To determine the  ability of intra- run variability (IRV) weighting to  consistently detect active voxels, we first acquired  fMRI data from a sample of healthy subjects, each of  whom performed 4 runs (4 blocks each) of self-paced  finger-tapping. Each subject's data was analyzed  using single-run general linear model (GLM), and  each block was then analyzed separately to calculate  the IRV weighting. Results show that integrating IRV  information into standard single-subject GLM  activation maps significantly improved the  reliability (p = 0.007) of the single-subject fMRI  data. This suggests that taking IRV into account can  help identify the most constant and relevant  neuronal activity at the single- subject level.}  }  @Article{bianciardi.EA:09,  author = {M. Bianciardi and L. Bianchi and G. Garreffa and  M. Abbafati and F. {Di Russo} and M.G. Marciani and  E. Macaluso},  title = {Single-epoch analysis of interleaved evoked  potentials and fMRI responses during steady-state  visual stimulation},  journal = ClinNeurophys,  year = 2009,  volume = 120,  pages = {738--747},  abstract = {Objective: Aim of the study was to record BOLD-fMRI  interleaved with evoked potentials for single-  epochs of visual stimulation and to investigate the  possible relationship between these two  measures. Methods: Sparse recording of fMRI and EEG  allowed us to measure BOLD responses and evoked  potentials on an epoch-by-epoch basis. To obtain  robust estimates of evoked potentials, we used  blocks of contrast- reversing visual stimuli  eliciting steady-state visual evoked potentials  (SSVEPs). For each block we acquired one volume of  fMRI data and we then tested for co-variations  between SSVEPs and fMRI signals. Our analyses tested  for frequency-specific co-variation between the two  measurements that could not be explained by the mere  presence/absence of the visual stimulation.  Results: Condition-specific single-epoch SSVEPs and  fMRI responses were observed at occipital  sites. Combined SSVEPs–fMRI analysis at the  single-epoch level did not reveal any significant  correlation between the two recordings. However,  both signals contained stimulation-specific linear  decreases that may relate to neuronal habituation.  Conclusions: Our findings demonstrate robust  estimation of single-epoch evoked potentials and  fMRI responses during interleaved recording, using  visual steady-state stimulation. Significance:  Single-epochs analysis of evoked potentials and fMRI  signals is feasible for interleaved SSVEPs–fMRI  recordings.}  }  @Article{bittar.EA:99,  author = {R.G. Bittar and M. Ptito and J. Faubert and  S.O. Dumoulin and A. Ptito},  title = {Activation of the Remaining Hemisphere Following  Stimulation of the Blind Hemifield in  Hemispherectomized Subjects},  journal = NI,  year = 1999,  volume = 10,  number = 3,  pages = {339--346},  abstract = {We used functional magnetic resonance imaging (fMRI)  to investigate the neural substrates mediating  residual vision in the “blind” hemifield of  hemispherectomized patients. The visual stimuli were  semicircular gratings moving in opposite directions  on a dynamic random-dot background. They were  specifically constructed to eliminate intra- and  extraocular light scatter and optimize the  activation of extrastriate cortical areas and their  subcortical relays. Multislice T2*-weighted gradient  echo (GE) echoplanar imaging (EPI) images (TR/TE = 4  s/45 ms, flip angle 90°) were acquired during  activation and baseline visual stimulation. An  activation minus baseline subtraction was performed,  and the acquired t statistic map transformed into  the stereotaxic coordinate space of Talairach and  Tournoux. In seven normal control subjects, right  hemifield stimulation produced significant  activation foci in contralateral V1/V2, V3/V3A, VP,  and V5 (MT). Significant activation was also  produced in homologous regions of the right  occipital lobe with left hemifield  stimulation. Stimulation of the intact hemifield in  hemispherectomized patients resulted in activation  of similar areas exclusively within the  contralateral hemisphere. Stimulation of the anopic  hemifield produced statistically significant  activation in the ipsilateral occipital lobe  (putative area V5 or MT) and areas V3/V3A in the  only subject with blindsight. We conclude that the  remaining hemisphere may contribute to residual  visual functions in the blind hemifield of  hemispherectomized patients, possibly through the  collicular–pulvinar route since the activated areas  are known to receive their afferents from these  subcortical nuclei.}  }  @Article{block:11,  author = {Ned Block},  title = {Perceptual consciousness overflows cognitive access},  journal = TIC,  year = 2011,  volume = 15,  number = 12,  pages = {567--575},  abstract = {One of the most important issues concerning the  foun- dations of conscious perception centers on the  question of whether perceptual consciousness is rich  or sparse. The overflow argument uses a form of  ‘iconic memory’ to argue that perceptual  consciousness is richer (i.e., has a higher  capacity) than cognitive access: when observ- ing a  complex scene we are conscious of more than we can  report or think about. Recently, the overflow argu-  ment has been challenged both empirically and  concep- tually. This paper reviews the controversy,  arguing that proponents of sparse perception are  committed to the postulation of (i) a peculiar kind  of generic conscious representation that has no  independent rationale and (ii) an unmotivated form  of unconscious representation that in some cases  conflicts with what we know about unconscious  representation. }  }  @Article{bordier.EA:15,  author = {C{\'e}cile Bordier and Jean-MichelHup{\'e} and  Michel Dojat},  title = {Quantitative evaluation of {fMRI} retinotopic maps,  from V1 to V4, for cognitive experiments},  journal = FHN,  year = 2015,  volume = 9,  number = 277,  note = {doi:10.3389/fnhum.2015.00277},  abstract = {FMRI retinotopic mapping is a non-invasive technique  for the delineation of low-level visual areas in  individual subjects. It generally relies upon the  analysis of functional responses to periodic visual  stimuli that encode eccentricity or polar angle in  the visual field. This technique is used in vision  research when the precise assignation of brain  activation to retinotopic areas is an issue. It  involves processing steps computed with different  algorithms and embedded in various software  suites. Manual intervention may be needed for some  steps. Although the diversity of the available  processing suites and manual interventions may  potentially introduce some differences in the final  delineation of visual areas, no documented  comparison between maps obtained with different  procedures has been reported in the literature. To  explore the effect of the processing steps on the  quality of the maps obtained, we used two tools,  BALC, which relies on a fully automated procedure,  and BrainVoyager, where areas are delineated “by  hand” on the brain surface. To focus on the mapping  procedures specifically, we used the same SPM  pipeline for pretreatment and the same tissue  segmentation tool. We document the consistency and  differences of the fMRI retinotopic maps obtained  from “routine retinotopy” experiments on 10  subjects. The maps obtained by skilled users are  never fully identical. However, the agreement  between the maps, around 80% for low-level areas, is  probably sufficient for most applications. Our  results also indicate that assigning cognitive  activations, following a specific experiment (here,  color perception), to individual retinotopic maps is  not free of errors. We provide measurements of this  error, that may help for the cautious interpretation  of cognitive activation projection onto fMRI  retinotopic maps. On average, the magnitude of the  error is about 20%, with much larger differences in  a few subjects. More variability may even be  expected with less trained users or using different  acquisition parameters and preprocessing chains.}  }  @Article{brett.EA:01,  author = {Matthew Brett and Alexander P. Leff and Chris Rorden  and John Ashburner},  title = {Spatial Normalization of Brain Images with Focal  Lesions Using Cost Function Masking},  journal = NI,  year = 2001,  volume = 14,  number = 2,  pages = {486--500},  abstract = {In studies of patients with focal brain lesions, it  is often useful to coregister an image of the  patient's brain to that of another subject or a  standard template. We refer to this process as  spatial normalization. Spatial normalization can  improve the presentation and analysis of lesion  location in neuropsychological studies; it can also  allow other data, for example from functional  imaging, to be compared to data from other patients  or normal controls. In functional imaging, the  standard procedure for spatial normalization is to  use an automated algorithm, which minimizes a  measure of difference between image and template,  based on image intensity values. These algorithms  usually optimize both linear (translations,  rotations, zooms, and shears) and nonlinear  transforms. In the presence of a focal lesion,  automated algorithms attempt to reduce image  mismatch between template and image at the site of  the lesion. This can lead to significant  inappropriate image distortion, especially when  nonlinear transforms are used. One solution is to  use cost-function masking—masking the areas used in  the calculation of image difference—to exclude the  area of the lesion, so that the lesion does not bias  the transformations. We introduce and evaluate this  technique using normalizations of a selection of  brains with focal lesions and normal brains with  simulated lesions. Our results suggest that  cost-function masking is superior to the standard  approach to this problem, which is affine-only  normalization; we propose that cost-function masking  should be used routinely for normalizations of  brains with focal lesions.}  }  @Article{bridge.EA:08,  author = {Holly Bridge and Owen Thomas and Saâd Jbabdi and  Alan Cowey},  title = {Changes in connectivity after visual cortical brain  damage underlie altered visual function},  journal = {Brain},  year = 2008,  volume = 131,  number = 6,  pages = {1433--1444},  abstract = {The full extent of the brain's ability to compensate  for damage or changed experience is yet to be  established. One question particularly important for  evaluating and understanding rehabilitation  following brain damage is whether recovery involves  new and aberrant neural connections or whether any  change in function is due to the functional  recruitment of existing pathways, or  both. Blindsight, a condition in which subjects with  complete destruction of part of striate cortex (V1)  retain extensive visual capacities within the  clinically blind field, is an excellent example of  altered visual function. Since the main pathway to  the visual cortex is destroyed, the spared or  recovered visual ability must arise from either an  existing alternative pathway, or the formation of a  new pathway. Using diffusion-weighted MRI, we show  that both controls and blindsight subject GY, whose  left V1 is destroyed, show an ipsilateral pathway  between LGN (lateral geniculate nucleus) and human  motion area MT+/V5 (bypassing V1). However, in  addition, GY shows two major features absent in  controls: (i) a contralateral pathway from right LGN  to left MT+/V5, (ii) a substantial cortico-cortical  connection between MT+/V5 bilaterally. Both  observations are consistent with previous functional  MRI data from GY showing enhanced ipsilateral  activation in MT+/V5. There is also evidence for a  pathway in GY from left LGN to right MT+/V5,  although the lesion makes its quantification  difficult. This suggests that employing alternative  brain regions for processing of information  following cortical damage in childhood may  strengthen or establish specific connections.}  }  @Article{bridge.EA:10,  author = {Holly Bridge and Stephen L. Hicks and Jingyi Xie and  Thomas W. Okell and Sabira Mannan and Iona Alexander  and Alan Cowey and Christopher Kennard},  title = {Visual activation of extra-striate cortex in the  absence of {V1} activation},  journal = {Neuropsychologia},  year = 2010,  volume = 48,  pages = {4148--4154},  abstract = {When the primary visual cortex (V1) is damaged,  there are a number of alternative pathways that can  carry visual information from the eyes to  extrastriate visual areas. Damage to the visual  cortex from trauma or infarct is often unilateral,  extensive and includes gray matter and white matter  tracts, which can disrupt other routes to residual  visual function. We report an unusual young patient,  SBR, who has bilateral damage to the gray matter of  V1, sparing the adjacent white matter and  surrounding visual areas. Using functional magnetic  resonance imaging (fMRI), we show that area MT+/V5  is activated bilaterally to visual stimulation,  while no significant activity could be measured in  V1. Additionally, the white matter tracts between  the lateral geniculate nucleus (LGN) and V1 appear  to show some degeneration, while the tracts between  LGN and MT+/V5 do not differ from  controls. Furthermore, the bilateral nature of the  damage suggests that residual visual capacity does  not result from strengthened interhemispheric  connections. The very specific lesion in SBR  suggests that the ipsilateral connection between LGN  and MT+/V5 may be important for residual visual  function in the presence of damage to V1.}  }  @Article{brown.roosgilbert:95,  author = {Tracy L. Brown and Linda Roos-Gilbert},  title = {Automaticity and word perception: Evidence from  Stroop and Stroop dilution effects},  journal = JEPlmc,  year = 1995,  volume = 95,  number = 21,  pages = 1395,  annote = {DON'T HAVE PAPER},  abstract = {Examines the theory that word perception is  automatic. Role of the automatic nature of word  perception in theories of reading skill; Stroop and  Stroop dilution effects as a source of support for  the automaticity of word perception; Alternative  explanations of Stroop dilution; Attention capture;  Presentation of neutral words at fixation; Early  visual interference account.}  }  @Article{bullier:01,  author = {Jean Bullier},  title = {Integrated model of visual processing},  journal = BRR,  year = 2001,  volume = 36,  number = {2--3},  pages = {96--107},  abstract = {Cortical processing of visual information requires  that information be exchanged between neurons coding  for distant regions in the visual field. It is  argued that feedback connections are the best  candidates for such rapid long-distance  interconnections. In the integrated model,  information arriving in the cortex from the  magnocellular layers of the lateral geniculate  nucleus is first sent and processed in the parietal  cortex that is very rapidly activated by a visual  stimulus. Results from this first-pass computation  are then sent back by feedback connections to areas  V1 and V2 that act as ‘active black-boards’ for the  rest of the visual cortical areas: information  retroinjected from the parietal cortex is used to  guide further processing of parvocellular and  koniocellular information in the inferotemporal  cortex.}  }  @Article{calderone.EA:13,  author = {Daniel J. Calderone and Antígona Martinez and Vance  Zemon and Matthew J. Hoptman and George Hu and Jade  E. Watkins and Daniel C. Javitt and Pamela  D. Butler},  title = {Comparison of psychophysical, electrophysiological,  and {fMRI} assessment of visual contrast responses  in patients with schizophrenia},  journal = NI,  year = 2013,  volume = 67,  pages = {153--162},  abstract = {Perception has been identified by the NIMH-sponsored  Cognitive Neuroscience Treatment Research to Improve  Cognition in Schizophrenia (CNTRICS) group as a  useful domain for assessing cognitive deficits in  patients with schizophrenia. Specific measures of  contrast gain derived from recordings of  steady-state visual evoked potentials (ssVEP) have  demonstrated neural deficits within the visual  pathways of patients with  schizophrenia. Psychophysical measures of contrast  sensitivity have also shown functional loss in these  pa- tients. In the current study, functional  magnetic resonance imaging (fMRI) was used in  conjunction with ssVEP and contrast sensitivity  testing to elucidate the neural underpinnings of  these deficits. During fMRI scanning, participants  viewed 1) the same low and higher spatial frequency  stimuli used in the psychophys- ical contrast  sensitivity task, at both individual detection  threshold contrast and at a high contrast; and 2)  the same stimuli used in the ssVEP paradigm, which  were designed to be biased toward either the  magnocellular or parvocellular visual  pathway. Patients showed significant impairment in  contrast sensitivity at both spatial frequencies in  the psychophysical task, but showed reduced  occipital activation volume for low, but not higher,  spatial frequency at the low and high contrasts  tested in the magnet. As expected, patients  exhibited selective deficits under the  magnocellular-biased ssVEP condition. However,  occipital lobe fMRI responses demonstrated the same  general pattern for magnocellular- and  parvocellular-biased stimuli across groups. These  results indicate dissociation between the fMRI  measures and the psychophysical/ssVEP  measures. These latter measures appear to have  greater value for the functional assessment of the  contrast deficits ex- plored here.}  }  @Article{celeghin.EA:15RT,  author = {Alessia Celeghin and Marissa Barabas and Francesca  Mancini and Matteo Bendini and Emilio Pedrotti and  Massimo Prior and Anna Cantagallo and Silvia Savazzi  and Carlo A. Marzi},  title = {Speeded manual responses to unseen visual stimuli in  hemianopic patients: What kind of blindsight?},  journal = ConscCog,  year = 2015,  volume = 32,  pages = {6--14},  abstract = {Blindsight, i.e., unconscious visually guided  behaviour triggered by stimuli presented to a  cortically blind hemifield, has been typically found  either by using direct (forced choice) or indirect  (interhemispheric) methods. However, one would  expect to find blindsight also in fast responses to  suddenly appearing visual stimuli, a reminiscence of  evolutionary ancient adaptive behaviour. In this  study we provide preliminary evidence of this form  of blindsight by using a conservative method for  assessing blindsight based on a comparison between  the cumulative probability functions (CPFs) of  simple reaction times to blind and intact field  stimuli. Furthermore, in two patients with  blindsight we provided evidence that their  above-chance unconscious responses were likely to be  triggered by the intact hemisphere.}  }  @Article{celeghin.EA:15,  author = {Alessia Celeghin and Silvia Savazzi and Marissa  Barabas and Matteo Bendini and Carlo A. Marzi},  title = {Blindsight is sensitive to stimulus numerosity and  configuration: evidence from the redundant signal  effect},  journal = EBR,  year = 2015,  volume = 233,  number = 5,  pages = {1617--1623},  annote = {One important, yet relatively unexplored ques- tion  is whether blindsight, i.e., unconscious visually  guided behavior in hemianopic patients, is endowed  with basic perceptual properties such as detecting  stimulus numerosity and overall  configuration. Rather than a forced-choice pro-  cedure in which patients are supposed to guess about  stim- uli presented to the blind hemifield, we used  a redundant signal effect paradigm, i.e., the  speeding of simple reaction time (RT) when  presenting multiple versus single similar  stimuli. The presence of an effect of numerosity for  the (unseen) stimuli presented to the blind field  was indirectly assessed by measuring RT to bilateral  versus unilateral stimuli presented to the intact  hemifield. Chronic hemian- opic patients were tested  with unilateral or bilateral black dots, both of  which could be either single or quadruple. The  latter could either have a fixed spatial  configuration repre- senting a diamond or be  randomly spatially assembled on every trial. Both  configurations covered the same extent of visual  field and had the overall same luminance. We found  that a numerosity effect as a result of increasing  the num- ber of stimuli in the blind field was  indeed present but only with the diamond  configuration. This is a convincing evi- dence that  this form of blindsight does not depend upon  stimulus numerosity per se but is likely to be  related to the presence of structured and memorized  rather than meaning- less changing stimuli.}  }  @Article{chawla.EA:99,  author = {D. Chawla and G. Rees and K. J. Friston},  title = {The physiological basis of attentional modulation in  extrastriate visual areas},  journal = NatNeuro,  year = 1999,  volume = 2,  number = 7,  pages = {671--676},  abstract = {Selective attention to color or motion enhances  activity in specialized areas of extrastriate  cortex, but mechanisms of attentional modulation  remain unclear. By dissociating modulation of  visually evoked transient activity from the baseline  for a particular attentional set, human functional  neuroimaging was used to investigate the  physiological basis of such effects. Baseline  activity in motion- and color-sensitive areas of  extrastriate cortex was enhanced by selective  attention to these attributes, even without moving  or colored stimuli. Further, visually evoked  responses increased along with baseline  activity. These results are consistent with the  hypothesis that attention modulates sensitivity of  neuronal populations to inputs by changing  background activity.}  }  @Article{chen.cave:13,  author = {Zhe Chen and Kyle R. Cave},  title = {Perceptual load vs. dilution: the roles of  attentional focus, stimulus category, and target  predictability},  journal = FP,  year = 2013,  volume = 4,  pages = 327,  abstract = {Many studies have shown that increasing the number  of neutral stimuli in a display decreases distractor  interference. This result has been interpreted  within two different frameworks; a perceptual load  account, based on a reduction in spare resources,  and a dilution account, based on a degradation in  distractor representation and/or an increase in  crosstalk between the distractor and the neutral  stimuli that contain visually similar features. In  four experiments, we systematically manipulated the  extent of attentional focus, stimulus category, and  preknowledge of the target to examine how these  factors would interact with the display set size to  influence the degree of distractor processing.  Display set size did not affect the degree of  distractor processing in all situations. Increasing  the number of neutral items decreased distractor  processing only when a task induced a broad  attentional focus that included the neutral stimuli,  when the neutral stimuli were in the same category  as the target and distractor, and when the  preknowledge of the target was insufficient to guide  attention to the target efficiently. These results  suggest that the effect of neutral stimuli on the  degree of distractor processing is more complex than  previously assumed. They provide new insight into  the competitive interactions between bottom-up and  top-down processes that govern the efficiency of  visual selective attention. }  }  @Article{clapp.gazzaley:12,  author = {Wesley C. Clapp and Adam Gazzaley},  title = {Distinct mechanisms for the impact of distraction  and interruption on working memory in aging},  journal = {Neurobiology of Aging},  year = 2012,  volume = 33,  pages = {134--148},  abstract = {Interference is known to negatively impact the  ability to maintain information in working memory  (WM), an effect that is exacerbated with  aging. Here, we explore how distinct sources of  interference, i.e., distraction (stimuli  to-be-ignored) and interruption (stimuli requiring  attention), differentially influence WM in younger  and older adults. EEG was recorded while  participants engaged in three versions of a delayed-  recognition task: no interference, a distracting  stimulus, and an interrupting stimulus presented  during WM maintenance. Behaviorally, both types of  interference negatively impacted WM accuracy in  older adults significantly more than younger adults  (with a larger deficit for interruptions). N170  latency measures revealed that the degree of  processing both distractors and interruptors  predicted WM accuracy in both populations. However,  while WM impairments could be explained by excessive  attention to distractors by older adults (a  suppression deficit), impairment induced by  interruption were not clearly mediated by  age-related increases in attention to  interruptors. These results suggest that distinct  underlying mechanisms mediate the impact of  different types of external interference on WM in  normal aging.}  }  @Article{cocchi.EA:11,  author = {Luca Cocchi and Ulrike Toepel and Marzia {De Lucia}  and Roberto Martuzzi and Stephen J. Wood and Olivia  Carter and Micah M. Murray},  title = {Working memory load improves early stages of  independent visual processing},  journal = {Neuropsychologia},  year = 2011,  volume = 49,  number = 1,  pages = {92--102},  abstract = {Increasing evidence suggests that working memory and  perceptual processes are dynamically interrelated  due to modulating activity in overlapping brain  networks. However, the direct influence of working  memory on the spatio-temporal brain dynamics of  behaviorally relevant intervening information  remains unclear. To investigate this issue, subjects  performed a visual proximity grid perception task  under three different visual--spatial working memory  (VSWM) load conditions. VSWM load was manipulated by  asking subjects to memorize the spatial locations of  6 or 3 disks. The grid was always presented between  the encoding and recognition of the disk pattern. As  a baseline condition, grid stimuli were presented  without a VSWM context. VSWM load altered both  perceptual performance and neural networks active  during intervening grid encoding. Participants  performed faster and more accurately on a  challenging perceptual task under high VSWM load as  compared to the low load and the baseline  condition. Visual evoked potential (VEP) analyses  identified changes in the configuration of the  underlying sources in one particular period  occurring 160--190 ms post-stimulus onset. Source  analyses further showed an occipito-parietal  down-regulation concurrent to the increased  involvement of temporal and frontal resources in the  high VSWM context. Together, these data suggest that  cognitive control mechanisms supporting working  memory may selectively enhance concurrent visual  processing related to an independent goal. More  broadly, our findings are in line with theoretical  models implicating the engagement of frontal regions  in synchronizing and optimizing mnemonic and  perceptual resources towards multiple goals.}  }  @Article{corballis:09correx,  author = {Michael C. Corballis},  title = {Comparing a single case with a control sample:  Correction and further comment},  journal = {Neuropsychologia},  year = 2009,  volume = 47,  number = 13,  pages = {2696--2697},  abstract = {Crawford, Garthwaite and Howell [Crawford, J. R.,  Garthwaite, P. H., & Howell, D. C. On comparing a  single case with a control sample: An alternative  perspective. Neuropsychologia, in press] correctly  point out that the t-test I suggested for asking  whether a single case belongs in the population from  which a control sample is drawn is incorrect. In  fact, the corrected formula can be interpreted  either as addressing this question, or as asking  whether the single case and the control sample are  drawn from populations with the same mean. Extension  to analysis of variance is recommended for factorial  experiments, where single case and control sample  are tested under different experimental  conditions. It is not recommended for studies in  which different tests are administered and treated  as different levels of the same dependent variable,  as in tests for neuropsychological dissociations.}  }  @Article{corballis:09,  author = {Michael C. Corballis},  title = {Comparing a single case with a control sample:  Correction and further comment},  journal = {Neuropsychologia},  year = 2009,  volume = 47,  number = 13,  pages = {2687--2689},  abstract = {Crawford and Garthwaite [Crawford, J. R. &  Garthwaite, P. H. (2002). Investigation of the  single case in neuropsychology: Confidence limits on  the abnormality and test score  differences. Neuropsychologia, 40, 1196–1208] have  proposed an adjusted t-test, widely used in  experimental neuropsychology, for comparing a single  case with a control sample. This test does not  assess whether the single-case score belongs in the  population from which the control sample is drawn,  but rather whether the mean of the distribution from  which the case was drawn differs significantly from  the mean of the control population. This approach is  readily extended to more complex designs in which  the analysis of variance is appropriate, and the  single case is treated as belonging to a group of  size one. The main qualification in using either  this or Crawford and Howell's approach is that it  makes the untestable assumption of homogeneity of  variance between the two populations, but a simple  adjustment either to the t-test or to the analysis  of variance allows one to draw conclusions about the  relation of the case itself to the control  population.}  }  @Article{cox.savoy:03,  author = {David D Cox and Robert L Savoy},  title = {Functional magnetic resonance imaging ({fMRI})  ``brain reading'': detecting and classifying  distributed patterns of fMRI activity in human  visual cortex},  journal = NI,  year = 2003,  volume = 19,  number = 2,  pages = {261--270},  abstract = {Traditional (univariate) analysis of functional MRI  (fMRI) data relies exclusively on the information  contained in the time course of individual  voxels. Multivariate analyses can take advantage of  the information contained in activity patterns  across space, from multiple voxels. Such analyses  have the potential to greatly expand the amount of  information extracted from fMRI data sets. In the  present study, multivariate statistical pattern  recognition methods, including linear discriminant  analysis and support vector machines, were used to  classify patterns of fMRI activation evoked by the  visual presentation of various categories of  objects. Classifiers were trained using data from  voxels in predefined regions of interest during a  subset of trials for each subject  individually. Classification of subsequently  collected fMRI data was attempted according to the  similarity of activation patterns to prior training  examples. Classification was done using only small  amounts of data (20 s worth) at a time, so such a  technique could, in principle, be used to extract  information about a subject’s percept on a near  real-time basis. Classifiers trained on data  acquired during one session were equally accurate in  classifying data collected within the same session  and across sessions separated by more than a week,  in the same subject. Although the highest  classification accuracies were obtained using  patterns of activity including lower visual areas as  input, classification accuracies well above chance  were achieved using regions of interest restricted  to higher-order object-selective visual areas. In  contrast to typical fMRI data analysis, in which  hours of data across many subjects are averaged to  reveal slight differences in activation, the use of  pattern recognition methods allows a subtle 10-way  discrimination to be performed on an essentially  trial-by-trial basis within individuals,  demonstrating that fMRI data contain far more  information than is typically appreciated.}  }  @Article{crawford.EA:03dissoc,  author = {John R. Crawford and Paul H. Garthwaite and Colin  D. Gray},  title = {Wanted: fully operational definitions of  dissociations in single--case studies},  journal = {Cortex},  year = 2003,  volume = 39,  pages = {357--370},  abstract = {In contrast to the careful consideration given to  the issue of what we can infer from dissociations in  single-case studies, the more basic question of how  we decide whether a dissociation is present has been  relatively neglected. Proposals are made for fully  operational definitions of a deficit, classical and  strong dissociations, and double dissociations. In  developing these definitions it was assumed that  they should be based on the use of inferential  rather than descriptive statistical methods. The  scope of these definitions is limited to typical  single-case studies in which patients are compared  to control samples of a modest size. The operational  definition of a classical dissociation incorporates  a requirement that a patient’s performance on Task X  is significantly different from Task Y, in addition  to the “standard” requirement that the patient has a  deficit on Task X and is within normal limits on  Task Y. We ran a simulation to estimate the Type I  error rates when the criteria for dissociations are  applied and found these to be low (Type I errors  were defined as identifying an individual from the  control population as having a dissociation). The  inferential methods for testing whether the various  criteria are met make use of t-distributions. These  methods are contrasted with the widespread use of z  to test for a deficit or a difference between  tasks. In the latter approach the statistics of the  control sample are treated as parameters; this is  not appropriate when, as is normally the case, the  control sample size is modest in size.}  }  @Article{crawford.EA:03,  author = {John R. Crawford and Paul H. Garthwaite and David  C. Howell and Annalena Venneri},  title = {Intra-individual measures of association in  neuropsychology: Inferential methods for comparing a  single case with a control or normative sample},  journal = JINS,  year = 2003,  volume = 9,  pages = {989-1000},  abstract = {Performance on some neuropsychological tests is best  expressed as an intra-individual measure of  association (such as a parametric or non-parametric  correlation coefficient or the slope of a regression  line). Examples of the use of intra-individual  measures of association (IIMAs) include the  quantification of performance on tests designed to  assess temporal order memory or the accuracy of time  estimation. The present paper presents methods for  comparing a patient’s performance with a control or  normative sample when performance is expressed as an  IIMA. The methods test if there is a significant  difference between a patient’s IIMA and those  obtained from controls, yield an estimate of the  abnormality of the patient’s IIMA, and provide  confidence limits on the level of abnormality. The  methods can be used with normative or control  samples of any size and will therefore be of  particular relevance to single-case researchers. A  method for comparing the difference between a  patient’s scores on two measures with the  differences observed in controls is also described  (one or both measures can be IIMAs). All the methods  require only summary statistics (rather than the raw  data from the normative or control sample); it is  hoped that this feature will encourage the  development of norms for tasks that use IIMAs to  quantify performance. Worked examples of the  statistical methods are provided using data from a  clinical case and controls. A computer program (for  PCs) that implements the methods is described and  made available.}  }  @Article{crawford.garthwaite:05,  author = {John R. Crawford and Paul H. Garthwaite},  title = {Testing for Suspected Impairments and Dissociations  in Single-Case Studies in Neuropsychology:  Evaluation of Alternatives Using {Monte Carlo}  Simulations and Revised Tests for Dissociations},  journal = Neuropsychology,  year = 2005,  volume = 19,  number = 3,  pages = {318--331},  abstract = {In neuropsychological single-case studies, a patient  is compared with a small control sample. Methods of  testing for a deficit on Task X, or a significant  difference between Tasks X and Y, either treat the  control sample statistics as parameters (using z and  zD) or use modified t tests. Monte Carlo simulations  demonstrated that if z is used to test for a  deficit, the Type I error rate is high for small  control samples, whereas control of the error rate  is essentially perfect for a modified t  test. Simulations on tests for differences revealed  that error rates were very high for zD. A new method  of testing for a difference (the revised  standardized difference test) achieved good control  of the error rate, even with very small sample  sizes. A computer program that implements this new  test (and applies criteria to test for classical and  strong dissociations) is made available.}  }  @Article{crawford.garthwaite:07,  author = {John R. Crawford and Paul H. Garthwaite},  title = {Comparison of a single case to a control or  normative sample in neuropsychology: Development of  a {Bayesian} approach},  journal = CogNeuropsych,  year = 2007,  volume = 24,  number = 4,  pages = {343--372},  abstract = {Frequentist methods are available for comparison of  a patient’s test score (or score difference) to a  control or normative sample; these methods also  provide a point estimate of the percentage of the  population that would obtain a more extreme score  (or score difference) and, for some problems, an  accompanying interval estimate (i.e., confidence  limits) on this percentage. In the present paper we  develop a Bayesian approach to these  problems. Despite the very different approaches, the  Bayesian and frequentist methods yield equivalent  point and interval estimates when (a) a case’s score  is compared to that of a control sample, and (b)  when the raw (i.e., unstandardized) difference  between a case’s scores on two tasks are compared to  the differences in controls. In contrast, the two  approaches differ with regard to point estimates of  the abnormality of the difference between a case’s  standardized scores. The Bayesian method for  standardized differences has the advantages that (a)  it can directly evaluate the probability that a  control will obtain a more extreme difference score,  (b) it appropriately incorporates error in  estimating the standard deviations of the tasks from  which the patient’s difference score is derived, and  (c) it provides a credible interval for the  abnormality of the difference between an  individual’s standardized scores; this latter  problem has failed to succumb to frequentist  methods. Computer programs that implement the  Bayesian methods are described and made available.}  }  @Article{crawford.EA:09,  author = {John R. Crawford and Paul H. Garthwaite and David  C. Howell},  title = {On comparing a single case with a control sample: An  alternative perspective},  journal = {Neuropsychologia},  year = 2009,  volume = 47,  number = 13,  pages = {2690--2695},  abstract = {Corballis [Corballis, M. C. (2009). Comparing a  single case with a control sample: Refinements and  extensions. Neuropsychologia] offers an interesting  position paper on statistical inference in  single-case studies. The following points arise: (1)  Testing whether we can reject the null hypothesis  that a patient's score is an observation from the  population of control scores can be a legitimate aim  for single-case researchers, not just  clinicians. (2) Counter to the claim made by  Corballis [Corballis, M. C. (2009). Comparing a  single case with a control sample: Refinements and  extensions. Neuropsychologia], Crawford and Howell's  [Crawford, J. R., & Howell, D. C. (1998). Comparing  an individual's test score against norms derived  from small samples. The Clinical Neuropsychologist,  12, 482–486] method does test whether we can reject  the above null hypothesis. (3) In all but the most  unusual of circumstances Crawford and Howell's  method can also safely be used to test whether the  mean of a notional patient population is lower than  that of a control population, should  neuropsychologists wish to construe the test in this  way. (4) In contrast, the method proposed by  Corballis is not legitimate for either of these  purposes because it fails to allow for uncertainty  over the control mean (as a result Type I errors  will not be under control). (5) The use of a mixed  ANOVA design to compare a case to controls (with or  without the adjustment proposed by Corballis) is  beset with problems but these can be overcome using  alternative methods.}  }  @Article{crawford.EA:10,  author = {John R. Crawford and Paul H. Garthwaite and Sara  Porter},  title = {Point and interval estimates of effect sizes for the  case-controls design in neuropsychology: Rationale,  methods, implementations, and proposed reporting  standards},  journal = CogNeuropsych,  year = 2010,  volume = 27,  number = 3,  pages = {245--260},  abstract = {It is increasingly common for group studies in  neuropsychology to report effect sizes. In contrast  this is rarely done in single-case studies (at least  in those studies that employ a case-controls  design). The present paper sets out the advantages  of reporting effect sizes, derives suitable effect  size indexes for use in single-case studies, and  develops methods of supplementing point estimates of  effect sizes with interval estimates. Computer  programs that implement existing classical and  Bayesian inferential methods for the single case (as  developed by Crawford, Garthwaite, Howell, and col-  leagues) are upgraded to provide these point and  interval estimates. The upgraded programs can be  downloaded from  www.abdn.ac.uk/~psy086/dept/Single_Case_Effect_Sizes.htm}  }  @Article{crawford.EA:11,  author = {John R. Crawford and Paul H. Garthwaite and Kevin  Ryan},  title = {Comparing a single case to a control sample: Testing  for neuropsychological deficits and dissociations in  the presence of covariates},  journal = {Cortex},  year = 2011,  volume = 47,  pages = {1166--1178},  abstract = {Existing inferential methods of testing for a  deficit or dissociation in the single case are  extended to allow researchers to control for the  effects of covariates. The new (Bayesian) methods  provide a significance test, point and interval  estimates of the effect size for the difference  between the case and controls, and point and  interval estimates of the abnor- mality of a case’s  score, or standardized score difference. The methods  have a wide range of potential applications, e.g.,  they can provide a means of increasing the  statistical power to detect deficits or  dissociations, or can be used to test whether  differences between a case and controls survive  partialling out the effects of potential confounding  variables. The methods are implemented in a series  of computer programs for PCs (these can be down-  loaded from  www.abdn.ac.uk/wpsy086/dept/Single_Case_Covariates.htm). Illustrative  examples of the methods are provided.}  }  @article{crinion.EA:07,  author = {Jenny Crinion and John Ashburner and Alex Leff and  Matthew Brett and Cathy Price and Karl Friston},  title = {Spatial normalization of lesioned brains:  Performance evaluation and impact on {fMRI}  analyses},  journal = NI,  year = 2007,  volume = 37,  number = 3,  pages = {866--875},  abstract = {A key component of group analyses of neuroimaging  data is precise and valid spatial normalization  (i.e., inter-subject image registration). When  patients have structural brain lesions, such as a  stroke, this process can be confounded by the lack  of correspondence between the subject and  standardized template images. Current procedures for  dealing with this problem include regularizing the  estimate of warping parameters used to match  lesioned brains to the template, or “cost function  masking”; both these solutions have significant  drawbacks. We report three experiments that identify  the best spatial normalization for structurally  damaged brains and establish whether differences  among normalizations have a significant effect on  inferences about functional activations. Our novel  protocols evaluate the effects of different  normalization solutions and can be applied easily to  any neuroimaging study. This has important  implications for users of both structural and  functional imaging techniques in the study of  patients with structural brain damage.}  }  @Article{danckert.EA:03,  author = {James Danckert and Patrice Revol and Laure Pisella  and Pierre Krolak-Salmon and Alain Vighetto and  Melvyn A. Goodale and Yves Rossetti},  title = {Measuring unconscious actions in action-blindsight:  exploring the kinematics of pointing movements to  targets in the blind field of two patients with  cortical hemianopia},  journal = {Neuropsychologia},  year = 2003,  volume = 41,  pages = {1068--1081},  abstract = { We tested two patients with posterior cerebral  lesions on two pointing tasks. In the first task,  the patients pointed to targets presented on a touch  screen monitor and pointing accuracy was  recorded. One patient (JR) demonstrated good  localisation of targets presented to her blind field  while the other patient (YP) did not. Movement  kinematics were measured in the second task to  compare the kinematics of movements made to sighted  field targets with those made to blind field  targets. For this version of the task both patients  demonstrated above chance localisation of blind  field targets although the slope of the relationship  between the end of pointing movements and the target  locations was significantly steeper for JR than for  YP. Furthermore, JR showed a kinematic profile for  movements made to blind field targets that mirrored  the profile of kinematics to sighted field  targets. That is, both peak velocity and time to  peak velocity increased with increasing target  eccentricity for movements made to blind and sighted  field targets alike. Although patient YP now showed  more reliable spatial localisation on this pointing  task when compared with the touch screen task, his  kinematics for movements made to targets in his  blind field were quite different from those made to  targets in his sighted field. Based on the patients’  CT scans, we suggest that the superior performance  of patient JR is a consequence of greater sparing of  her parietal cortex in the damaged hemisphere.}  }  @Article{danckert.rossetti:05,  author = {James Danckert and Yves Rossetti},  title = {Blindsight in action: what can the different  sub-types of blindsight tell us about the control of  visually guided actions?},  journal = NBR,  year = 2005,  volume = 29,  number = 7,  pages = {1035--1046},  abstract = {Blindsight broadly refers to the paradoxical  neurological condition where patients with a visual  field defect due to a cortical lesion nevertheless  demonstrate implicit residual visual sensitivity  within their field cut. The aim of this paper is  twofold. First, through a selective review of the  blindsight literature we propose a new taxonomy for  the subtypes of residual abilities described in  blindsight. Those patients able to accurately act  upon blind field stimuli (e.g. by pointing or  saccading towards them) are classified as having  ‘action-blindsight’, those whose residual functions  can be said to rely to some extent upon attentive  processing of blind field stimuli are classified as  demonstrating ‘attention-blindsight’, while finally,  patients who have somewhat accurate perceptual  judgements for blind field stimuli despite a  complete lack of any conscious percept, are  classified as having ‘agnosopsia’—literally meaning  ‘not knowing what one sees’. We also address the  possible neurological substrates of these residual  sensory processes. Our second aim was to  investigate the most striking subtype of blindsight,  action-blindsight. We review the data relevant to  this subtype and the hypotheses proposed to account  for it, before speculating on how action-blindsight  may inform our normal models of visuomotor control.}  }  @InCollection{danckert.mirsattarri:12,  author = {James Danckert and Seyed M. Mirsattarri},  title = {Neuroimaging of Single Cases: Benefits and Pitfalls},  booktitle = {Neuroimaging -- Cognitive and Clinical Neuroscience},  publisher = {InTech},  year = 2012,  editor = {Peter Bright},  note = {doi: 10.5772/22983}  }  @InCollection{dirusso.EA:02,  author = {Francesco {Di Russo} and Wolfgang  A. Teder-S{\"a}̈llej{\"a}rvi and Steven A. Hillyard},  title = {{Steady-State VEP} and Attentional Visual  Processing},  booktitle = {The Cognitive Electrophysiology of Mind and Brain},  publisher = {Academic Press},  year = 2002,  editor = {Alberto Zani and Alice Mado Proverbio},  pages = {259--274},  address = {Amsterdam},  abstract = {The vast majority of studies that investi- gated  attentional modulation of the visual evoked  potential (VEP) have been confined to the transient  responses evoked by iso- lated stimuli. This class  of potentials is evoked by stimuli having an  asynchronous and low repetition rate (not faster  than 2 stimuli per second). These potentials are  called “transient” because the slow rate of  stimulation allows the sensory pathways to recover  or “reset” before the next stimulus appears. When  visual stimuli are presented at a constant rate that  is rapid enough to prevent the evoked neural  activity from returning to base line state, the  elicited response becomes continuous and is called  the steady-state visual evoked potential  (SSVEP). With steady-state stimulation the typical  VEP wave form is markedly changed. For instance, the  transient VEP includes three major early components:  the C1 at 60–80 msec, the P1 at 80–120 msec, and the  N1 at 120–180 msec. (see Fig. 1). At more rapid  stimulation rates, the brain response to the same  stimulus becomes sinusoidal and is typically  modulated at the fundamental stimulus frequency in  the case of an unstructured stimulus (e.g. flash) or  at the second harmonic (double the stimulation  frequency) if the stimulus is a pattern-reversal  (Regan, 1989). Like any sinusoidal wave form, the  SSVEP can be measured in terms of its amplitude and  phase. The phase is a joint function of the stimulus  frequency and the time delay between stimulus and  brain response. The amplitude indicates the rela-  tive magnitude of a given harmonic of the response  and, as for transient evoked potentials, is measured  in microvolts. The amplitude and phase of the SSVEP  vary as function of the temporal frequency, spatial  frequency, contrast, luminance, and hue of the  driving stimulus (Regan, 1989).}  }  @Article{dilks.EA:07,  author = {Daniel D. Dilks and John T. Serences and Benjamin  J. Rosenau and Steven Yantis and Michael McCloskey},  title = {Human adult cortical reorganization and consequent  visual distortion},  journal = JoN,  year = 2007,  volume = 27,  number = 36,  pages = {9585--9594},  abstract = {Neural and behavioral evidence for cortical  reorganization in the adult somatosensory system  after loss of sensory input (e.g., amputation) has  been well documented. In contrast, evidence for  reorganization in the adult visual system is far  less clear: neural evidence is the subject of  controversy, behavioral evidence is sparse, and  studies combining neural and behavioral evidence  have not previously been reported. Here, we report  converging behavioral and neuroimaging evidence from  a stroke patient (B.L.) in support of cortical  reorganization in the adult human visual  system. B.L.'s stroke spared the primary visual  cortex (V1), but destroyed fibers that normally  provide input to V1 from the upper left visual field  (LVF). As a consequence, B.L. is blind in the upper  LVF, and exhibits distorted perception in the lower  LVF: stimuli appear vertically elongated, toward and  into the blind upper LVF. For example, a square  presented in the lower LVF is perceived as a  rectangle extending upward. We hypothesized that the  perceptual distortion was a consequence of cortical  reorganization in V1. Extensive behavioral testing  supported our hypothesis, and functional magnetic  resonance imaging (fMRI) confirmed V1  reorganization. Together, the behavioral and fMRI  data show that loss of input to V1 after a stroke  leads to cortical reorganization in the adult human  visual system, and provide the first evidence that  reorganization of the adult visual system affects  visual perception. These findings contribute to our  understanding of the human adult brain's capacity to  change and has implications for topics ranging from  learning to recovery from brain damage.}  }  @Article{dirusso.EA:03,  author = {Francesco {Di Russo} and Antigona Mart{\'i}nez and  Steven A. Hillyard},  title = {Source Analysis of Event-related Cortical Activity  during Visuo-spatial Attention},  journal = CC,  year = 2003,  volume = 13,  pages = {486--499},  abstract = {Recordings of event-related potentials (ERPs) were  combined with structural and functional magnetic  resonance imaging (fMRI) to study the  spatio-temporal patterns of cortical activity that  underlie visual–spatial attention. Small  checkerboard stimuli were flashed in random order to  the four quadrants of the visual field at a rapid  rate while subjects attended to stimuli in one  quadrant at a time. Attended stimuli elicited  enhanced ERP components in the latency range 80–200  ms that were co-localized with fMRI activations in  multiple extrastriate cortical regions. The earliest  ERP component (C1 at 50–90 ms) was unaffected by  attention and was localized by dipole modeling to  calcarine cortex. A longer latency deflection in the  150–225 ms range that was accounted for by this same  calcarine source, however, did show consistent  modulation with attention. This late attention  effect, like the C1, inverted in polarity for upper  versus lower field stimuli, consistent with a neural  generator in primary visual cortex (area V1). These  results provide support to current hypotheses that  spatial attention in humans is associated with  delayed feedback to area V1 from higher extrastriate  areas that may have the function of improving the  salience of stimuli at attended locations.}  }  @Article{dirusso.EA:07,  author = {Francesco {Di Russo} and Sabrina Pitzalis and Teresa  Aprile and Grazia Spitoni and Fabiana Patria and  Alessandra Stella and Donatella Spinelli and Steven  A. Hillyard},  title = {Spatiotemporal analysis of the cortical sources of  the steady-state visual evoked potential},  journal = HBM,  year = 2007,  volume = 28,  number = 4,  pages = {323--334},  abstract = {This study aimed to characterize the neural  generators of the steady-state visual evoked  potential (SSVEP) to repetitive, 6 Hz  pattern-reversal stimulation. Multichannel scalp  recordings of SSVEPs and dipole modeling techniques  were combined with functional magnetic resonance  imaging (fMRI) and retinotopic mapping in order to  estimate the locations of the cortical sources  giving rise to the SSVEP elicited by pattern  reversal. The time-varying SSVEP scalp topography  indicated contributions from two major cortical  sources, which were localized in the medial  occipital and mid-temporal regions of the  contralateral hemisphere. Colocalization of dipole  locations with fMRI activation sites indicated that  these two major sources of the SSVEP were located in  primary visual cortex (V1) and in the motion  sensitive (MT/V5) areas, respectively. Minor  contributions from mid-occipital (V3A) and ventral  occipital (V4/V8) areas were also  considered. Comparison of SSVEP phase information  with timing information collected in a previous  transient VEP study (Di Russo et al. [2005]  Neuroimage 24:874-886) suggested that the sequence  of cortical activation is similar for steady-state  and transient stimulation. These results provide a  detailed spatiotemporal profile of the cortical  origins of the SSVEP, which should enhance its use  as an efficient clinical tool for evaluating  visual-cortical dysfunction as well as an  investigative probe of the cortical mechanisms of  visual-perceptual processing.}  }  @Article{dirusso.EA:12,  author = {Francesco {Di Russo} and Alessandra Stella and  Grazia Spitoni and Francesca Strappini and Stefano  Sdoia and Gaspare Galati and Steven A. Hillyard and  Donatella Spinelli and Sabrina Pitzalis},  title = {Spatiotemporal Brain Mapping of Spatial Attention  Effects on Pattern-Reversal {ERPs}},  journal = HBM,  year = 2012,  volume = 33,  pages = {1334--1351},  abstract = {Recordings of event-related potentials (ERPs) were  combined with structural and functional magnetic  resonance imaging (fMRI) to investigate the timing  and localization of stimulus selection processes  during visual-spatial attention to pattern-reversing  gratings. Pattern reversals were presented in random  order to the left and right visual fields at a rapid  rate, while subjects attended to the rever- sals in  one field at a time. On separate runs, stimuli were  presented in the upper and lower visual  quadrants. The earliest ERP component (C1, peaking  at around 80 ms), which inverted in polarity for  upper versus lower field stimuli and was localized  in or near visual area V1, was not modulated by  attention. In the latency range 80–250 ms, multiple  components were elicited that were increased in  amplitude by attention and were colocalized with  fMRI activations in specific visual cortical  areas. The principal anatomical sources of these  attention-sensitive components were localized by  fMRI-seeded dipole modeling as follows: P1 (ca. 100  ms—source in motion-sensitive area MTþ), C2 (ca. 130  ms— same source as C1), N1a (ca. 145 ms—source in  horizontal intraparietal sulcus), N1b (ca. 165 ms—  source in fusiform gyrus, area V4/V8), N1c (ca. 180  ms—source in posterior intraparietal sulcus, area  V3A), and P2 (ca. 220 ms—multiple sources, including  parieto-occipital sulcus, area V6). These results  support the hypothesis that spatial attention acts  to amplify both feed-forward and feedback signals in  multiple visual areas of both the dorsal and ventral  streams of processing.}  }  @Article{dirusso.EA:16,  author = {Francesco {Di Russo} and Giuliana Lucci and  Valentina Sulpizio and Marika Berchicci and  Donatella Spinelli and Sabrina Pitzalis and Gaspare  Galati},  title = {Spatiotemporal brain mapping during preparation,  perception, and action},  journal = NI,  year = 2016,  volume = 126,  pages = {1--14},  abstract = {Deciding whether to act or not to act is a  fundamental cognitive function. To avoid incorrect  responses, both re- active and proactive modes of  control have been postulated. Little is known,  however, regarding the brain imple- mentation of  proactive mechanisms, which are deployed prior to an  actual need to inhibit a response. Via a combination  of electrophysiological and neuroimaging measures  (recorded in 21 and 16 participants, respective-  ly), we describe the brain localization and timing  of neural activity that underlies the anticipatory  proactive mechanism. From these results, we conclude  that proactive control originates in the inferior  Frontal gyrus, is established well before stimulus  perception, and is released concomitantly with  stimulus appearance. Stimulus perception triggers  early activity in the anterior insula and  intraparietal cortex contralateral to the responding  hand; these areas likely mediate the transition from  perception to action. The neural activities leading  to the de- cision to act or not to act are described  in the framework of a three-stage model that  includes perception, action, and anticipatory  functions taking place well before stimulus onset.}  }  @Article{driver.mattingley:98,  author = {Jon Driver and Jason B. Mattingley},  title = {Parietal neglect and visual awareness},  journal = NatNeuro,  year = 1998,  volume = 1,  pages = {17--22},  abstract = {The last decade has seen a resurgence of interest in  the neural correlates of conscious vision, with most  discussion focused on the 'blindsight' that can  follow damage to primary visual cortex, in the  occipital lobe. We suggest that new insights into  the neural basis of visual awareness may be gleaned  from a different neuropsychological phenomenon,  namely visual 'neglect' after injury to regions in  the parietal lobe. Neglect provides several  revealing contrasts with occipital blindsight. Here  we summarise four key findings. First, unlike the  deficits caused by damage to primary visual cortex,  the loss of awareness in parietal neglect is  characteristically not strictly retinotopic. Second,  visual segmentation processes are preserved in  neglect, and can influence what will reach the  patient's awareness. Third, extensive unconscious  processing takes place for those stimuli on the  neglected side which escape awareness, including  some degree of object identification. Finally,  parietal damage affects initial stages of motor  planning as well as perception. These findings are  consistent with recent data on single-cell activity  in the monkey brain. They also suggest why areas in  the inferior parietal lobe may play a prominent role  in visual awareness.}  }  @Article{eklund.EA:12,  author = {Anders Eklund and Mats Andersson and Camilla  Josephson and Magnus Johannesson and Hans Knutsson},  title = {Does parametric {fMRI} analysis with {SPM} yield  valid results? -- An empirical study of 1484 rest  datasets},  journal = NI,  year = 2012,  volume = 61,  number = 3,  pages = {565--578},  abstract = {The validity of parametric functional magnetic  resonance imaging (fMRI) analysis has only been  reported for simulated data. Recent advances in  computer science and data sharing make it possible  to analyze large amounts of real fMRI data. In this  study, 1484 rest datasets have been analyzed in  SPM8, to estimate true familywise error rates. For a  familywise significance threshold of 5\%, significant  activity was found in 1\%–70\% of the 1484 rest  datasets, depending on repetition time, paradigm and  parameter settings. This means that parametric  significance thresholds in SPM both can be  conservative or very liberal. The main reason for  the high familywise error rates seems to be that the  global AR(1) auto correlation correction in SPM  fails to model the spectra of the residuals,  especially for short repetition times. The findings  that are reported in this study cannot be  generalized to parametric fMRI analysis in general,  other software packages may give different  results. By using the computational power of the  graphics processing unit (GPU), the 1484 rest  datasets were also analyzed with a random  permutation test. Significant activity was then  found in 1\%–19\% of the datasets. These findings  speak to the need for a better model of temporal  correlations in fMRI timeseries.}  }  @InProceedings{eklund.EA:15,  author = {Anders Eklund and Thomas Nichols and Mats Andersson  and Hans Knutsson},  title = {Empirically investigating the statistical validity  of {SPM, FSL and AFNI} for single subject {fMRI}  analysis},  booktitle = {IEEE 12th International Symposium on Biomedical  Imaging (ISBI)},  year = 2015,  pages = {1376--1380},  organization = {IEEE},  abstract = {The software packages SPM, FSL and AFNI are the most  widely used packages for the analysis of functional  magnetic resonance imaging (fMRI) data. Despite this  fact, the validity of the statistical methods has  only been tested using simulated data. By analyzing  resting state fMRI data (which should not contain  specific forms of brain activity) from 396 healthy  con- trols, we here show that all three software  packages give in- flated false positive rates  (4\%--96\% compared to the expected 5\%). We isolate  the sources of these problems and find that SPM  mainly suffers from a too simple noise model, while  FSL underestimates the spatial smoothness. These  results highlight the need of validating the  statistical methods being used for fMRI.}  }  @Article{ellis.EA:06,  author = {Kathryn A. Ellis and Richard B. Silberstein and  Pradeep J. Nathan},  title = {Exploring the temporal dynamics of the spatial  working memory n-back task using steady state visual  evoked potentials {(SSVEP)}},  journal = NI,  year = 2006,  volume = 31,  pages = {1741--1751},  abstract = {The neural networks associated with spatial working  memory (SWM) are well established. However, the  temporal dynamics of SWM-related brain activity are  less clear. This study examined changes in temporal  neurophysiology during the spatial n-back task using  steady state probe topography (SSPT) to record  cortical steady state visual evoked potentials  (SSVEPs) at 64 scalp locations. Twenty healthy male  volunteers participated in the study. The findings  identified three different time periods of  significance during the spatial n-back task— an  early perceptual/encoding period (approximately  0–500 ms), an early delay period just following the  stimulus disappearing from view (approximately 850 –  1400 ms), and a late period lasting the final second  of the delay and anticipation of the new stimulus  (approxi- mately 2500 – 3500 ms). The delay period  was associated with increases in frontal and  occipital region amplitude, consistent with previous  findings in more basic working memory tasks. The two  different SSVEP components during the delay appear  reflective of the additional ‘‘executive’’ demands  associated with the n-back and may suggest variable  roles for the PFC during different stages of the  delay. All three n-back levels demonstrated a  relative consistent electrophysiological profile,  indicating that this pattern is specific to the  spatial n-back task. Nevertheless, these findings  supported the hypothesis that memory load modulates  activity within the networks identified, consistent  with previous neuroimaging studies. The current  findings may offer a framework in which to further  investigate the temporal aspects of SWM.}  }  @InCollection{fahle:03,  author = {Manfred Fahle},  title = {Failures of visual analysis: scotoma, agnosia, and  neglect},  booktitle = {The Neuropsychology of Vision},  publisher = {Oxford University Press},  year = 2003,  editor = {Manfred Fahle and Mark Greenlee},  pages = {179--258},  address = {Oxford},  abstract = {This chapter deals with different types of failure  to analyse the visual world, starting with the  complete loss of vision, blindness, as well as  blindness for circumscribed parts of the visual  field, scotomata. It also reports some syndromes  characterized by difficulties in perception of and  discrimination between different domains or  submodalities of vision, while leaving most others  intact, such as acquired inability to assess visual  motion, or acquired colour blindness caused by  cerebral damage (achromatopsia). It then provides an  overview of the most astonishing examples of failed  visual perception: simultanagnosia; neglect; and  Balint's syndrome. These syndromes share the  phenomenon that conscious perception of objects in  (parts of) the peripheral visual field is absent in  spite of intact stimulation of the primary visual  cortex by these objects. It also attempts to fill in  some gaps of knowledge by means of speculation to  produce a consistent taxonomy of symptoms  encountered in visual neuropsychology. It combines  the theoretical background with clinical findings  and with a description of at least some of the  methods used to diagnose patients suffering from  failures of visual analysis.}  }  @article{faulkner.foster:02,  author = {Deborah Faulkner and Jonathan Foster},  title = {The Decoupling of {"}Explicit{"}and {"}Implicit{"}  Processing in Neuropsychological Disorders: Insights  into the Neural Basis of Consciousness?},  journal = {Psyche -- An interdisciplinary journal of research  on consciousness},  year = 2002,  volume = 8,  number = 2,  abstract = {A key element of the distinction between explicit  and implicit cognitive functioning is the presence  or absence of conscious awareness. In this review,  we consider the proposal that neuropsychological  disorders can best be considered in terms of a  decoupling between preserved implicit or unconscious  processing and impaired explicit or conscious  processing. Evidence for dissociations between  implicit and explicit processes in blindsight,  amnesia, object agnosia, prosopagnosia,  hemi-neglect, and aphasia is examined. The  implications of these findings for a) our  understanding of a variety of neuropsychological  disorders, b) the conceptualization of normal  cognitive functioning, c) the neural basis of  consciousness, and d) the clinical rehabilitation of  brain-injured individuals are also discussed.}  }  @Article{fawcett.EA:04,  author = {Ian P. Fawcett and Gareth R. Barnes and Arjan  Hillebrand and Krish D. Singh},  title = {The temporal frequency tuning of human visual cortex  investigated using synthetic aperture magnetometry},  journal = NI,  year = 2004,  volume = 21,  number = 4,  pages = {1542--1555},  abstract = {Using synthetic aperture magnetometry (SAM) analyses  of magnetoencephalographic (MEG) data, we  investigated the variation in cortical response  magnitude and frequency as a function of stimulus  temporal frequency. In two separate experiments, a  reversing checkerboard stimulus was used in the  right or left lower visual field at frequencies from  0 to 21 Hz. Average temporal frequency tuning curves  were constructed for regions-of-interest located  within medial visual cortex and V5/MT. In medial  visual cortex, it was found that both the frequency  and magnitude of the steady-state response varied as  a function of the stimulus frequency, with multiple  harmonics of the stimulus frequency being found in  the response. The maximum fundamental response was  found at a stimulus frequency of 8 Hz, whilst the  maximum broadband response occurred at 4 Hz. In  contrast, the magnitude and frequency content of the  evoked onset response showed no dependency on  stimulus frequency. Whilst medial visual cortex  showed a power increase during stimulation,  extra-striate areas such as V5/MT exhibited a  bilateral event-related desynchronisation (ERD). The  frequency content of this ERD did not depend on the  stimulus frequency but was a broadband power  reduction across the 5–20 Hz frequency range. The  magnitude of this ERD within V5/MT was strongly  low-pass tuned for stimulus frequency, and showed  only a moderate preference for stimuli in the  contralateral visual field.}  }  @Article{fellrath.ptak:15,  author = {Julia Fellrath and Radek Ptak},  title = {The role of visual saliency for the allocation of  attention: Evidence from spatial neglect and  hemianopia},  journal = {Neuropsychologia},  year = 2015,  volume = 73,  pages = {70--81},  abstract = {Visual scanning and exploration of natural scenes  not only depends on specific objects, but also on  local features. Models of spatial attention propose  that features such as orientation or colour are  processed pre-attentively and in parallel. According  to these models attention interferes at a later  stage, where features are combined into a  representation of visual saliency. Saliency is a  good predictor of ocular fixations during scanning  of static pictures. Here, we tested whether  fixations of patients with left spatial neglect,  hemianopia, or both can be predicted based on local  image content. Participants were asked to freely  scan natural images while saccades and ocular  fixations were registered. Hemianopic patients  produced a similar distribution of fixations and  relied similarly on picture saliency as healthy  controls. In contrast, neglect patients looked to  image regions with increased saliency and higher  local orientation and intensity thresholds on the  neglected side of space. The reliance on increased  saliency during visual exploration was predicted by  damage to subcortical regions interconnecting the  inferior parietal and lateral premotor cortex. These  findings suggest that spatial neglect leads to a  combined attentive and pre-attentive deficit in the  processing of saliency and feature information.}  }  @Article{ffytche.zeki:11,  author = {Dominic H. Ffytche and Semir Zeki},  title = {The primary visual cortex, and feedback to it, are  not necessary for conscious vision},  journal = {Brain},  year = 2011,  volume = 134,  number = 1,  pages = {247--257},  abstract = {A compelling single case report of visual awareness  (visual qualia) without primary visual cortex would  be sufficient to refute the hypothesis that the  primary visual cortex and the back-projections to it  are necessary for conscious visual experience. In a  previous study, we emphasized the presence of crude  visual awareness in Patient G.Y., with a lesion of  the primary visual cortex, who is aware of, and able  to discriminate, fast-moving visual stimuli  presented to his blind field. The visual nature of  Patient G.Y.’s blind field experience has since been  questioned and it has been suggested that the  special circumstances of repeated testing over  decades may have altered Patient G.Y.’s visual  pathways. We therefore sought new evidence of visual  awareness without primary visual cortex in patients  for whom such considerations do not apply. Three  patients with hemianopic field defects (Patient  G.N. and Patient F.B. with MRI confirmed primary  visual cortex lesions, Patient C.G. with an inferred  lesion) underwent detailed psychophysical testing in  their blind fields. Visual stimuli were presented at  different velocities and contrasts in two- and  four-direction discrimination experiments and the  direction of motion and awareness reported using a  forced-choice paradigm. Detailed verbal reports were  also obtained of the nature of the blind field  experience with comparison of the drawings of the  stimulus presented in the blind and intact fields,  where possible. All three patients reported visual  awareness in their blind fields. Visual awareness  was significantly more likely when a moving stimulus  was present compared to no stimulus catch trials  (P < 0.01 for each subject). Psychophysical  performance in Patient F.B. and Patient G.N. was  consistent with the Riddoch syndrome, with higher  levels of visual awareness for moving compared to  static stimuli (P < 0.001) and intact direction  discrimination (P < 0.0001 for two- and  four-direction experiments). Although the blind  field experience of all three subjects was degraded,  it was clearly visual in nature. We conclude that  the primary visual cortex or back-projections to it  are not necessary for visual awareness.}  }  @Article{friston.EA:99,  author = {Karl J. Friston and Andrew P. Holmes and Keith  J. Worsley},  title = {How Many Subjects Constitute a Study?},  journal = NI,  year = 1999,  volume = 10,  pages = {1--5},  abstract = {In fMRI there are two classes of inference: one aims  to make a comment about the ‘‘typical’’  characteristics of a population, and the other about  ‘‘average’’ charac- teristics. The first pertains to  studies of normal sub- jects that try to identify  some qualitative aspect of normal functional  anatomy. The second class necessar- ily applies to  clinical neuroscience studies that want to make an  inference about quantitative differences of a  regionally specific nature. The first class of  inferences is adequately serviced by conjunction  analyses and fixed-effects models with relatively  small numbers of subjects. The second requires  random-effect analyses and larger cohorts.}  }  @Article{garcia.EA:13,  author = {Javier O. Garcia and Ramesh Srinivasan and John  T. Serences},  title = {Near-Real-Time Feature-Selective Modulations in  Human Cortex},  journal = CurrBio,  year = 2013,  volume = 23,  pages = {515--522},  abstract = {For neural activity to be linked with cognitive  function, infor- mation is needed about both the  temporal dynamics and the content of neural  codes. Traditionally, recording of single neurons in  animals has been the primary means of obtaining high  temporal resolution and precise information about  neural tuning properties such as selectivity for  different sensory features. Recent functional  magnetic resonance imaging (fMRI) studies in humans  have been able to measure feature selectivity within  specific subregions of sensory cortex (e.g.,  orientation selectivity in primary visual cortex, or  V1) [1, 2]. However, investigating the neural  mechanisms that support cognitive processing—which  often occurs rapidly on a subsecond scale—with a  temporally insensitive method such as fMRI severely  limits the types of inferences that can be  drawn. Here, we describe a new method for tracking  the rapid temporal evolution of feature-selective  information processing with scalp recordings of  electroen- cephalography (EEG). We generate  orientation-selective response profiles based on the  spatially distributed pattern of steady-state visual  evoked potential (SSVEP) responses to flickering  visual stimuli. Using this approach, we report a  multiplicative attentional modulation of these  feature- selective response profiles with a temporal  resolution of 24–120 ms, which is far faster than  that achieved with fMRI. Finally, we show that  behavioral performance on a discrimi- nation task  can be predicted based on the amplitude of these  temporally precise feature-selective response  profiles. This method thus provides a  high-temporal-resolution metric that can be used to  track the influence of cognitive manipu- lations on  feature-selective information processing in human  cortex.}  }  @Article{gazzaniga:95,  author = {Michael S. Gazzaniga},  title = {Principles of Human Brain Organization Derived from  Split-Brain Studies},  journal = {Neuron},  year = 1995,  volume = 14,  pages = {217--228}}  @Article{goebel.EA:01,  author = {Rainer Goebel and Lars Muckli and Friedhelm  E. Zanella and Wolf Singer and Petra Stoerig},  title = {Sustained extrastriate cortical activation without  visual awareness revealed by {fMRI} studies of  hemianopic patients},  journal = VisRes,  year = 2001,  volume = 41,  pages = {1459--1474},  abstract = {Patients with lesions in the primary visual cortex  (V1) may show processing of visual stimuli presented  in their field of cortical blindness even when they  report being unaware of the stimuli. To elucidate  the neuroanatomical basis of their residual visual  functions, we used functional magnetic resonance  imaging in two hemianopic patients, FS and GY. In  the first experiment, a rotating spiral stimulus was  used to assess the responsiveness of dorsal stream  areas. Although no response was detectable within  denervated or destroyed early visual cortex,  motion-sensitive areas (hMT+/V5) ipsilateral to the  lesion showed a strong sustained hemodynamic  response. In GY, this activation was at least as  strong as that of his contralesional hMT+/V5 to the  stimulus in the normal hemifield. In the second  experiment, coloured images of natural objects were  used to assess the responsiveness of ventral stream  areas. Again, no activity was detectable in  ipsilesional early visual areas, but extrastriate  areas in the lateral occipital cortex (hMT+/V5 and  LO) and within the posterior fusiform gyrus (V4/V8)  showed a robust sustained hemodynamic response. In  both experiments, we observed that ipsilesional  areas responded to stimuli presented in either  hemifield, whereas the normal hemisphere responded  preferentially to stimuli in the sighted  hemifield. As only one subject occasionally noticed  the onset of stimuluation in the impaired field, the  unexpectedly strong sustained activity in  ipsilesional dorsal and ventral cortical areas  appears to be insufficient to generate conscious  vision.}  }  @Article{gonzalezortiz.EA:13,  author = {S. Gonz{\'a}lez-Ortiz and L. Oleaga and T. Pujol and  S. Medrano and J. Rumi{\'a} and L. Caral and  T. Boget and J. Capellades and N. Bargall{\'}},  title = {Simple {fMRI} Postprocessing Suffices for Normal  Clinical Practice},  journal = AJNR,  year = 2013,  volume = 34,  pages = {1188--1193},  note = {dx.doi.org/10.3174/ajnr.A3381},  abstract = {BACKGROUND AND PURPOSE: Whereas fMRI postprocessing  tools used in research are accurate but unwieldy,  those used for clinical practice are user-friendly  but are less accurate. We aimed to determine whether  commercial software for fMRI postprocessing is  accurate enough for clinical practice. METHODS: Ten  volunteers underwent fMRI while performing motor and  language tasks (hand, foot, and orolingual  movements; verbal fluency; semantic judgment; and  oral comprehension). We compared visual concordance,  image quality (noise), voxel size, and radiologist  preference for the activation maps obtained by using  Neuro3D software (provided with our MR imaging  scanner) and by using the SPM program commonly used  in research. RESULTS: Maps obtained with the 2  methods were classified as "partially overlapping"  for 70\% for motor and 72\% for language paradigm  experiments and as “overlapping” in 30\% of motor  and in 15\% of language paradigm experiments.  CONCLUSIONS: fMRI is a helpful and robust tool in  clinical practice for planning neurosurgery. Widely  available commercial fMRI software can provide  reliable information for therapeutic management, so  sophisticated, less widely available software is  unnecessary in most cases.}  }  @Article{gorgolewski.EA:12,  author = {Krzysztof J. Gorgolewski and Amos J. Storkey and  Mark E. Bastin and Ian Whittle and Cyril Pernet},  title = {Single subject {fMRI} test–retest reliability  metrics and confounding factors},  journal = NI,  year = 2012,  volume = 69,  pages = {231--243},  abstract = {For the past twenty years, the tool of choice for  non-invasive study of human mind/brain relationships  has been functional Magnetic Res- onance Imaging  (fMRI). Despite the fact that it has been used in  thou- sands of studies, many of which have been  independently replicated, there is as yet no  consensus on how reliable fMRI measurements are  (Bennett and Miller, 2010). At the same time it is  widely accepted that fMRI can provide valuable  insights into the human brain even when used on the  single subject level. In other words, the result of  analysing fMRI time-series is not random. However,  it is also accepted that there is some variability  in the results that cannot be accounted for by  experimental variables. Understanding this  variability of fMRI is crucial to delineating limits  of fMRI as a research tool. The pursuit of  scientific truth is not the only motivation behind  understanding the reliability of fMRI. Shortly after  its inception fMRI was adapted for clinical use. For  example, presurgical mapping for ⁎ Corresponding  author at: Neuroinformatics and Computational  Neuroscience, Doctoral Training Centre, School of  Informatics, University of Edinburgh, 10 Crichton  Street, Edinburgh, EH8 9AB, UK. E-mail address:  [email protected] (K.J. Gorgolewski).  1053-8119. While the fMRI test–retest reliability has  been mainly investigated from the point of view of  group level studies, here we present analyses and  results for single-subject test–retest  reliability. One important aspect of group level  reliability is that not only does it depend on  between-session variance (test–retest), but also on  between-subject variance. This has partly led to a  debate regarding which reliability metric to use and  how different sources of noise contribute to  between-session variance. Focusing on single subject  reliability allows considering between-session  only. In this study, we measured test–retest  reliability in four behavioural tasks (motor  mapping, covert verb generation, overt word  repetition, and a landmark identification task) to  ensure generalisation of the results and at three  levels of data processing (time-series correlation,  t value variance, and overlap of thresholded maps)  to understand how each step influences the other and  how confounding factors influence reliability at  each of these steps. The contributions of  confounding factors (scanner noise, subject motion,  and coregistration) were investigated using multiple  regression and relative importance analyses at each  step. Finally, to achieve a fuller picture of what  constitutes a reliable task, we introduced a  bootstrap technique of within- vs. between-subject  variance. Our results show that (i) scanner noise  and coregistration errors have little contribution  to between-session variance (ii) subject motion  (especially correlat- ed with the stimuli) can have  detrimental effects on reliability (iii) different  tasks lead to different reliability re- sults. This  suggests that between-session variance in fMRI is  mostly caused by the variability of underlying  cognitive processes and motion correlated with the  stimuli rather than technical limitations of data  processing.}  }  @Article{gorgolewski.EA:13,  author = {Krzysztof J. Gorgolewski and Amos J. Storkey and  Mark E. Bastin and Ian Whittle and Cyril Pernet},  title = {Single subject {fMRI} test–retest reliability  metrics and confounding factors},  journal = NI,  year = 2013,  volume = 69,  pages = {231--243},  abstract = {While the fMRI test–retest reliability has been  mainly investigated from the point of view of group  level studies, here we present analyses and results  for single-subject test–retest reliability. One  important aspect of group level reliability is that  not only does it depend on between-session variance  (test–retest), but also on between-subject  variance. This has partly led to a debate regarding  which reliability metric to use and how different  sources of noise contribute to between-session  variance. Focusing on single subject reliability  allows considering between-session only. In this  study, we measured test–retest reliability in four  behavioural tasks (motor mapping, covert verb  generation, overt word repetition, and a landmark  identification task) to ensure generalisation of the  results and at three levels of data processing  (time-series correlation, t value variance, and  overlap of thresholded maps) to understand how each  step influences the other and how confounding  factors influence reliability at each of these  steps. The contributions of confounding factors  (scanner noise, subject motion, and coregistration)  were investigated using multiple regression and  relative importance analyses at each step. Finally,  to achieve a fuller picture of what constitutes a  reliable task, we introduced a bootstrap technique  of within- vs. between-subject variance. Our results  show that (i) scanner noise and coregistration  errors have little contribution to between-session  variance (ii) subject motion (especially correlat-  ed with the stimuli) can have detrimental effects on  reliability (iii) different tasks lead to different  reliability re- sults. This suggests that  between-session variance in fMRI is mostly caused by  the variability of underlying cognitive processes  and motion correlated with the stimuli rather than  technical limitations of data processing.}  }  @Book{greenstein.greenstein:00,  author = {Ben Greenstein and Adam Greenstein},  title = {Color Atlas of Neuroscience: Neuroanatomy and  Neurophysiology},  publisher = {Thieme},  year = 2000,  address = {Stuttgart}  }  @Article{grgic.EA:16,  author = {Regina Gregori Grgi{\vc} and Enrico Calore and  Claudio {de'Sperati}},  title = {Covert enaction at work: Recording the continuous  movements of visuospatial attention to visible or  imagined targets by means of Steady-State Visual  Evoked Potentials {(SSVEPs)}},  journal = {Cortex},  year = 2016,  volume = 74,  pages = {31--52},  annote = {Whereas overt visuospatial attention is customarily  measured with eye tracking, covert attention is  assessed by various methods. Here we exploited  Steady-State Visual Evoked Potentials (SSVEPs) e the  oscillatory responses of the visual cortex to  incoming flickering stimuli e to record the  movements of covert visuospatial attention in a way  operatively similar to eye tracking (attention  tracking), which allowed us to compare motion obser-  vation and motion extrapolation with and without eye  movements. Observers fixated a central dot and  covertly tracked a target oscillating horizontally  and sinusoidally. In the background, the left and  the right halves of the screen flickered at two  different fre- quencies, generating two SSVEPs in  occipital regions whose size varied reciprocally as  observers attended to the moving target. The two  signals were combined into a single quantity that  was modulated at the target frequency in a  quasi-sinusoidal way, often clearly visible in  single trials. The modulation continued almost  unchanged when the target was switched off and  observers mentally extrapolated its motion in  imagery, and also when observers pointed their  finger at the moving target during covert tracking,  or imagined doing so. The amplitude of modulation  during covert tracking was ~25--30\% of that  measured when observers followed the target with  their eyes. We used 4 electrodes in  parieto-occipital areas, but similar results were  achieved with a single electrode in Oz. In a second  experiment we tested ramp and step motion. During  overt tracking, SSVEPs were remarkably accurate,  showing both saccadic-like and smooth pursuit-like  modulations of cortical responsiveness, although  during covert tracking the modulation  deteriorated. Covert tracking was better with  sinusoidal motion than ramp motion, and better with  moving targets than stationary ones. The clear  modulation of cortical responsiveness recorded  during both overt and covert tracking, identical for  motion observation and motion extrapolation,  suggests to include covert attention movements in  enactive theories of mental imagery.}  }  @Article{haak.EA:14,  author = {Koen V. Haak and Dave R.M. Langers and Remco Renken  and Pim {van Dijk} and Johannes Borgstein and Frans  W. Cornelissen},  title = {Abnormal visual field maps in human cortex: A  mini-review and a case report},  journal = {Cortex},  year = 2014,  volume = 56,  pages = {14--25},  abstract = {Human visual cortex contains maps of the visual  field. Much research has been dedicated to answering  whether and when these visual field maps change if  critical components of the visual circuitry are  damaged. Here, we first provide a focused  mini-review of the functional magnetic resonance  imaging (fMRI) studies that have evaluated the human  cortical visual field maps in the face of retinal  lesions, brain injury, and atypical retinocortical  projections. We find that there is a fair body of  research that has found abnormal fMRI activity, but  also that this abnormal activity does not  necessarily stem from cortical remapping. The  abnormal fMRI activity can often be explained in  terms of task effects and/or the uncovering of  normally hidden system dynamics. We then present the  case of a 16-year-old patient who lost the entire  left cerebral hemisphere at age three for treatment  of chronic focal en- cephalitis (Rasmussen syndrome)  and intractable epilepsy. Using an fMRI retinotopic  mapping procedure and population receptive field  (pRF) modeling, we found that (1) despite the long  period since the hemispherectomy, the retinotopic  organization of early visual cortex remained  unaffected by the removal of an entire cerebral  hemisphere, and (2) the intact lateral occipital  cortex contained an exceptionally large  representation of the center of the visual  field. The same method also indicates that the  neuronal receptive fields in these lateral occipital  brain regions are extraordinarily small. These  features are clearly abnormal, but again they do not  necessarily stem from cortical remapping. For  example, the abnormal features can also be explained  by the notion that the hemispherectomy took place  during a critical period in the development of the  lateral occipital cortex and therefore arrested its  normal development. Thus, caution should be  exercised when interpreting abnormal fMRI activity  as a marker of cortical remapping; there are often  other explanations.}  }  @Book{haberland:07,  author = {Catherine Haberland},  title = {Clinical Neuropathology: text and color atlas},  publisher = {Demos Medical Publishing},  year = 2007,  address = {New York, USA}  }  @Article{hadjikhani.EA:01,  author = {Nouchine Hadjikhani and Margarita {Sanchez del Rio}  and Ona Wu and Denis Schwartz and Dick Bakker and  Bruce Fischl and Kenneth K. Kwong and F. Michael  Cutrer and Bruce R. Rosen and Roger B. H. Tootell  and A. Gregory Sorensen and Michael A. Moskowitz},  title = {Mechanisms of migraine aura revealed by functional  {MRI} in human visual cortex},  journal = PNAS,  year = 2001,  volume = 98,  number = 8,  pages = {4687--4692},  abstract = {Cortical spreading depression (CSD) has been  suggested to under- lie migraine visual  aura. However, it has been challenging to test this  hypothesis in human cerebral cortex. Using  high-field func- tional MRI with near-continuous  recording during visual aura in three subjects, we  observed blood oxygenation level-dependent (BOLD)  signal changes that demonstrated at least eight  character- istics of CSD, time-locked to percept  onset of the aura. Initially, a focal increase in  BOLD signal (possibly reflecting vasodilation),  developed within extrastriate cortex (area  V3A). This BOLD change progressed contiguously and  slowly (3.5 1.1 mm min) over occipital cortex,  congruent with the retinotopy of the visual per-  cept. Following the same retinotopic progression,  the BOLD signal then diminished (possibly reflecting  vasoconstriction after the initial vasodilation), as  did the BOLD response to visual activation. During  periods with no visual stimulation, but while the  subject was experiencing scintillations, BOLD signal  followed the retino- topic progression of the visual  percept. These data strongly sug- gest that an  electrophysiological event such as CSD generates the  aura in human visual cortex.}  }  @Article{henriksson.EA:07,  author = {L. Henriksson and A. Raninen and R. N{\"a}s{\"a}nen  and L. Hyv{\"a}rinen and S. Vanni},  title = {Training-induced cortical representation of a  hemianopic hemifield},  journal = JNNP,  year = 2007,  volume = 78,  pages = {74--81},  abstract = {Background: Patients with homonymous hemianopia  often have some residual sensitivity for visual  stimuli in their blind hemifield. Previous imaging  studies suggest an important role for extrastriate  cortical areas in such residual vision, but results  of training to improve vision in patients with  hemianopia are conflicting. Objective: To show that  intensive training with flicker stimulation in the  chronic stage of stroke can reorganise visual  cortices of an adult patient. Methods: A  61-year-old patient with homonymous hemianopia was  trained with flicker stimulation, starting 22 months  after stroke. Changes in functioning during training  were documented with magnetoencephalography, and the  cortical organisation after training was examined  with functional magnetic resonance imaging (fMRI).  Results: Both imaging methods showed that, after  training, visual information from both hemifields  was processed mainly in the intact hemisphere. The  fMRI mapping results showed the representations of  both the blind and the normal hemifield in the same  set of cortical areas in the intact hemisphere, more  specifically in the visual motion-sensitive area V5,  in a region around the superior temporal sulcus and  in retinotopic visual areas V1 (primary visual  cortex), V2, V3 and V3a. Conclusions: Intensive  training of a blind hemifield can induce cortical  reorganisation in an adult patient, and this case  shows an ipsilateral representation of the trained  visual hemifield in several cortical areas,  including the primary visual cortex. }  }  @Article{hirsch.EA:00,  author = {Joy Hirsch and Maximilian I. Ruge and Karl H.S. Kim  and Denise D. Correa and Jonathan D. Victor and  Norman R. Relkin and Douglas R. Labar and George  Krol and Mark H. Bilsky and Mark M. Souweidane and  Lisa M. DeAngelis},  title = {An Integrated Functional Magnetic Resonance Imaging  Procedure for Preoperative Mapping of Cortical Areas  Associated with Tactile, Motor, Language, and Visual  Functions},  journal = {Neurosurgery},  year = 2000,  volume = 47,  pages = {711--722},  abstract = {OBJECTIVE: To evaluate an integrated battery of  preoperative functional magnetic resonance imaging  (fMRI) tasks developed to identify cortical areas  associated with tactile, motor, language, and visual  functions. METHODS: Sensitivity of each task was  determined by the probability that a targeted region  was activated for both healthy volunteers (n 63) and  surgical patients with lesions in these critical  areas (n 125). Accuracy of each task was determined  by the correspondence between the fMRI maps and  intraoperative electrophysiological measurements,  including somatosensory evoked potentials (n 16),  direct cortical stimulation (n 9), and language  mapping (n 5), and by preoperative Wada tests (n 13)  and visual field examinations (n 6). RESULTS: For  healthy volunteers, the overall sensitivity was  100\% for identification of the central sulcus,  visual cortex, and putative Wernicke’s area, and  93\% for the putative Broca’s area (dominant  hemisphere). For patients with tumors affecting  these regions of interest, task sensitivity was 97\%  for identification of the central sulcus, 100\% for  the visual cortex, 91\% for the putative Wernicke’s  area, and 77\% for the putative Broca’s area. These  sensitivities were enhanced by the use of multiple  tasks to target related functions. Concordance of  the fMRI maps and intraoperative  electrophysiological measurements was observed  whenever both techniques yielded maps and Wada and  visual field examinations were consistent with fMRI  results. CONCLUSION: This integrated fMRI task  battery offers standardized and noninvasive  preoperative maps of multiple critical functions to  facilitate assessment of surgical risk, planning of  surgical routes, and direction of conventional,  intraoperative electrophysiological  procedures. Thus, a greater range of structural and  functional relationships is brought to bear in the  service of optimal outcomes for neurosurgery.}  }  @Article{huisman.EA:06,  author = {Thierry A. G. M. Huisman and Thomas Loenneker and  Gerd Barta and Matthias E. Bellemann and Juergen  Hennig and Joachim E. Fischer and Kamil A. Il'yasov},  title = {Quantitative diffusion tensor MR imaging of the  brain: field strength related variance of apparent  diffusion coefficient {(ADC)} and fractional  anisotropy {(FA)} scalars},  journal = ER,  year = 2006,  volume = 16,  number = 8,  pages = {1651--1658},  abstract = {The objectives were to study the "impact" of the  magnetic field strength on diffusion tensor imaging  (DTI) metrics and also to determine whether  magnetic-field-related differences in T2-relaxation  times of brain tissue influence DTI  measurements. DTI was performed on 12 healthy  volunteers at 1.5 and 3.0 Tesla (within 2 h) using  identical DTI scan parameters. Apparent diffusion  coefficient (ADC) and fractional anisotropy (FA)  values were measured at multiple gray and white  matter locations. ADC and FA values were compared  and analyzed for statistically significant  differences. In addition, DTI measurements were  performed at different echo times (TE) for both  field strengths. ADC values for gray and white  matter were statistically significantly lower at 3.0  Tesla compared with 1.5 Tesla (% change between  -1.94% and -9.79%). FA values were statistically  significantly higher at 3.0 Tesla compared with 1.5  Tesla (% change between +4.04 and 11.15%). ADC and  FA values are not significantly different for TE=91  ms and TE=125 ms. Thus, ADC and FA values vary with  the used field strength. Comparative clinical  studies using ADC or FA values should consequently  compare ADC or FA results with normative ADC or FA  values that have been determined for the field  strength used.}  }  @Article{iacoboni.zaidel:04,  author = {Marco Iacoboni and Eran Zaidel},  title = {Interhemispheric visuo-motor integration in humans:  the role of the superior parietal cortex},  journal = {Neuropsychologia},  year = 2004,  volume = 42,  number = 4,  pages = {419--425},  abstract = {We used event-related functional magnetic resonance  imaging (fMRI) to investigate the neural correlates  of basic interhemispheric visuo-motor  integration. In a simple reaction time task,  subjects responded to lateralized left and right  light flashes with unimanual left and right hand  responses. Typically, reaction times are faster for  uncrossed responses (that is, visual stimulus and  response hand on the same side) than for crossed  responses (that is, visual stimulus and response  hand on opposite sides). The chronometric difference  between crossed and uncrossed responses is called  crossed–uncrossed difference (CUD) and it is  typically taken to represent a behavioral estimate  of interhemispheric transfer time. The fMRI results  obtained in normal right-handers show that the  crossed conditions yielded greater activity,  compared to the uncrossed conditions, in bilateral  prefrontal, bilateral dorsal premotor, and right  superior parietal areas. These results suggest that  multiple transfers between the hemispheres occur in  parallel at the functional levels of sensory–motor  integration (posterior parietal), decision-making  (prefrontal) and preparation of motor response  (premotor). To test the behavioral significance of  these multiple transfers, we correlated the  individual CUDs with the difference in signal  intensity between crossed and uncrossed responses in  the prefrontal, dorsal premotor, and right superior  parietal activated areas. The analyses demonstrated  a strong correlation between the CUD and signal  intensity difference between crossed and uncrossed  responses in the right superior parietal  cortex. These data suggest a critical role of the  superior parietal cortex in interhemispheric  visuo-motor integration.}  }  @InCollection{iacoboni:05,  author = {Marco Iacoboni},  title = {Divided Attention in the Normal and the Split Brain:  Chronometry and Imaging},  booktitle = {Neurobiology of Attention},  publisher = {Academic Press},  year = 2005,  editor = {Laurent Itti and Geraint Rees and John K. Tsotsos},  pages = {363--367},  address = {San Diego, California, US},  abstract = {Divided attention is the ability to integrate in  parallel multiple stimuli. A relevant experimental  effect that has been studied for almost a century is  the redundant target effect. When multiple copies of  the same stimulus are presented to subjects, in  choice, go no-go, and even a simple reaction time  task, reaction times (RT) tend to be faster,  compared to RT to a single copy of the  stimulus. Paradoxically, this effect is larger in  split-brain patients when two stimuli are presented  in the two opposite hemifields. Recent RT and  imaging studies reviewed in this chapter suggest  that cortico-subcortical interactions between the  superior colliculus and the extrastriate cortex,  that are modulated by the corpus callosum, are  reflected in different levels of activation in  dorsal premotor cortex during divided attention  tasks and can account for the paradoxical  facilitation observed in split-brain patients.}  }  @Article{itti.EA:98,  author = {Laurent Itti and Christof Koch and Ernst Niebur},  title = {A Model of Saliency-Based Visual Attention for Rapid  Scene Analysis},  journal = {IEEE Transactions on Pattern Analysis and Machine  Intelligence,},  year = 1998,  volume = 20,  number = 11,  pages = {1254--1259},  abstract = {A visual attention system, inspired by the behavior  and the neuronal architecture of the early primate  visual system, is presented. Multiscale image  features are combined into a single topographical  saliency map. A dynamical neural network then  selects attended locations in order of decreasing  saliency. The system breaks down the complex problem  of scene understanding by rapidly selecting, in a  computationally efficient manner, conspicuous  locations to be analyzed in detail.}  }  @Book{jacobson.marcus:11,  author = {Stanley Jacobson and Elliott M. Marcus},  title = {Neuroanatomy  for the Neuroscientist},  publisher = {Springer},  year = 2011,  address = {New York, USA},  edition = {Second}}  @Article{james.EA:03,  author = {Thomas W. James and Jody Culham and G. Keith  Humphrey and A. David Milner and Melvyn A. Goodale},  title = {Ventral occipital lesions impair object recognition  but not object‐directed grasping: an {fMRI} study},  journal = {Brain},  year = 2003,  volume = 126,  pages = {2463--2475},  abstract = {D.F., a patient with severe visual form agnosia, has  been the subject of extensive research during the  past decade. The fact that she could process visual  input accurately for the purposes of guiding action  despite being unable to perform visual  discriminations on the same visual input inspired a  novel interpretation of the functions of the two  main cortical visual pathways or ‘streams’. Within  this theoretical context, the authors proposed that  D.F. had suffered severe bilateral damage to her  occipitotemporal visual system (the ‘ventral  stream’), while retaining the use of her  occipitoparietal visual system (the ‘dorsal  stream’). The present paper reports a direct test of  this idea, which was initially derived from purely  behavioural data, before the advent of modern  functional neuroimaging. We used functional MRI to  examine activation in her ventral and dorsal streams  during object recognition and object‐directed  grasping tasks. We found that D.F. showed no  difference in activation when presented with line  drawings of common objects compared with scrambled  line drawings in the lateral occipital cortex (LO)  of the ventral stream, an area that responded  differentially to these stimuli in healthy  individuals. Moreover, high‐resolution anatomical  MRI showed that her lesion corresponded bilaterally  with the location of LO in healthy participants. The  lack of activation with line drawings in  D.F. mirrors her poor performance in identifying the  objects depicted in the drawings. With coloured and  greyscale pictures, stimuli that she can identify  more often, D.F. did show some ventral‐stream  activation. These activations were, however, more  widely distributed than those seen in control  participants and did not include LO. In contrast to  the absent or abnormal activation observed during  these perceptual tasks, D.F. showed robust  activation in the expected dorsal stream regions  during object grasping, despite considerable atrophy  in some regions of the parietal lobes. In  particular, an area in the anterior intraparietal  sulcus was activated more for grasping an object  than for just reaching to that object, for both  D.F. and controls. In conclusion, we have been able  to confirm directly that D.F.’s visual form agnosia  is associated with extensive damage to the ventral  stream, and that her spared visuomotor skills are  associated with visual processing in the dorsal  stream.}  }  @Article{jamison.EA:15,  author = {Keith W. Jamison and Abhrajeet V. Roy and Sheng He  and Stephen A. Engel and Bin He},  title = {{SSVEP} signatures of binocular rivalry during  simultaneous {EEG and fMRI}},  journal = JNMeth,  year = 2015,  volume = 243,  pages = {53--62},  abstract = {Binocular rivalry is a perceptual phenomenon that  arises when two incompatible images are presented  separately, one to each eye, and the observer  experiences involuntary perceptual alternations  between the two images. If the two images are  flickering at two distinct frequencies,  electroencephalography (EEG) can be used to track  the frequency-tagged steady-state visually evoked  potential (SSVEP) driven by each image as they  compete for awareness, providing an objective  measure of the subjective perceptual state. This  spontaneous alternation in perceptual dominance is  believed to be driven by neural processes across  widespread regions in the brain, but the real-time  mechanisms of these processes remain unclear.}  }  @Article{janz.EA:00,  author = {Clemens Janz and Christina Schmitt and Oliver Speck  and J{\"u}rgen Hennig},  title = {Comparison of the hemodynamic response to different  visual stimuli in single-event and block stimulation  {fMRI} experiments},  journal = JMRI,  year = 2000,  volume = 12,  pages = {708--714},  abstract = {Experiments with three different types of basic  visual stimulation were performed to compare  cortical activation in single-event and block  trials. Independent of the stimulation paradigm, the  single-event presentation leads to highly consistent  signal responses regarding both the activated  cortical areas and the dynamics of the signal time  course. In contrast, signal time courses during  block paradigms depend on the stimulus applied and  are a complex and nonlinear function of the  single-event responses. Additionally, the initial  dip during the first 2 seconds after stimulus onset  is consistently observed. However, the small  amplitude change (−0.1% to −0.3%) requires signal  averaging to establish statistical significance of  the effect. Furthermore, different patterns of  activation were observed within the primary visual  cortex. In an anterior part of the primary visual  cortex, activation was only observed at the onset  and at the cessation of stimulation involving  luminance changes.}  }  @Article{jenkins.EA:03,  author = {Rob Jenkins and Nilli Lavie and Jon Driver},  title = {Ignoring famous faces: Category-specific dilution of  distractor interference},  journal = PP,  year = 2003,  volume = 65,  number = 2,  pages = {298--309},  abstract = {The extent to which famous distractor faces can be  ignored was assessed in six experiments. Subjects  categorized famous printed target names as those of  pop stars or politicians, while attempting to ignore  a flanking famous face distractor that could be  congruent (e.g., a politician's name and face) or  incongruent (e.g., a politician's name with a pop  star's face). Congruency effects on reaction times  indicated distractor intrusion. An additional,  response-neutral flanker (neither pop star nor  politician) could also be present. Congruency  effects from the critical distractor face were  reduced (diluted) by the presence of an intact  anonymous face, but not by phase-shifted versions,  inverted faces, or meaningful nonface objects. By  contrast, congruency effects from other types of  distracting objects (musical instruments, fruits),  when printed names for these classes were  categorized, were diluted equivalently by intact  faces, phase-shifted faces, or meaningful nonface  objects. Our results suggest that distractor faces  act differently from other types of distractors,  suffering from only face-specific capacity limits.  }  }  @Article{jones.cercignani:10,  author = {Derek K. Jones and Mara Cercignani},  title = {Twenty-five pitfalls in the analysis of diffusion  {MRI} data},  journal = NMR,  year = 2010,  volume = 23,  number = 7,  pages = {803--820},  abstract = {Obtaining reliable data and drawing meaningful and  robust inferences from diffusion MRI can be  challenging and is subject to many pitfalls. The  process of quantifying diffusion indices and  eventually comparing them between groups of subjects  and/or correlating them with other parameters starts  at the acquisition of the raw data, followed by a  long pipeline of image processing steps. Each one of  these steps is susceptible to sources of bias, which  may not only limit the accuracy and precision, but  can lead to substantial errors. This article  provides a detailed review of the steps along the  analysis pipeline and their associated  pitfalls. These are grouped into 1 pre-processing of  data; 2 estimation of the tensor; 3 derivation of  voxelwise quantitative parameters; 4 strategies for  extracting quantitative parameters; and finally 5  intra-subject and inter-subject comparison,  including region of interest, histogram,  tract-specific and voxel-based analyses. The article  covers important aspects of diffusion MRI analysis,  such as motion correction, susceptibility and eddy  current distortion correction, model fitting, region  of interest placement, histogram and voxel-based  analysis. We have assembled 25 pitfalls (several  previously unreported) into a single article, which  should serve as a useful reference for those  embarking on new diffusion MRI-based studies, and as  a check for those who may already be running studies  but may have overlooked some important  confounds. While some of these problems are well  known to diffusion experts, they might not be to  other researchers wishing to undertake a clinical  study based on diffusion MRI.}  }  @Article{kavcic.EA:15,  author = {Voyko Kavcic and Regina L. Triplett and Anasuya Das  and Tim Martin and Krystel R. Huxlin},  title = {Role of inter-hemispheric transfer in generating  visual evoked potentials in {V1}-damaged brain  hemispheres},  journal = {Neuropsychologia},  year = 2015,  volume = 68,  pages = {82--93},  abstract = {Partial cortical blindness is a visual deficit  caused by unilateral damage to the primary visual  cortex, a condition previously considered beyond  hopes of rehabilitation. However, recent data  demonstrate that patients may recover both simple  and global motion discrimination following intensive  training in their blind field. The present  experiments characterized motion-induced neural  activity of cortically blind (CB) subjects prior to  the onset of visual rehabilitation. This was done to  provide information about visual processing  capabilities available to mediate training-induced  visual improvements. Visual Evoked Po- tentials  (VEPs) were recorded from two experimental groups  consisting of 9 CB subjects and 9 age- matched,  visually-intact controls. VEPs were collected  following lateralized stimulus presentation to each  of the 4 visual field quadrants. VEP waveforms were  examined for both stimulus-onset (SO) and motion-  onset (MO) related components in postero-lateral  electrodes. While stimulus presentation to intact  re- gions of the visual field elicited normal SO-P1,  SO-N1, SO-P2 and MO-N2 amplitudes and latencies in  contralateral brain regions of CB subjects, these  components were not observed contralateral to  stimulus presentation in blind quadrants of the  visual field. In damaged brain hemispheres, SO-VEPs  were only recorded following stimulus presentation  to intact visual field quadrants, via  inter-hemispheric transfer. MO-VEPs were only  recorded from damaged left brain hemispheres,  possibly reflecting a native left/right asymmetry in  inter-hemispheric connections. The present findings  suggest that damaged brain hemi- spheres contain  areas capable of responding to visual  stimulation. However, in the absence of training or  rehabilitation, these areas only generate detectable  VEPs in response to stimulation of the intact hemi-  field of vision.}  }  @Article{kentridge.EA:99,  author = {R. W. Kentridge and C. A. Heywood and L. Weiskrantz},  title = {Attention without awareness in blindsight},  journal = PhilTransLonBio,  year = 1999,  volume = 266,  pages = {1805--1811},  abstract = {The act of attending has frequently been equated  with visual awareness. We examined this relationship  in `blindsight'öa condition in which the latter is  absent or diminished as a result of damage to the  primary visual cortex. Spatially selective visual  attention is demonstrated when information that  stimuli are likely to appear at a speci¢c location  enhances the speed or accuracy of detection of  stimuli subsequently presented at that location. In  a blindsight subject, we showed that attention can  confer an advantage in processing stimuli presented  at an attended location, without those stimuli  entering consciousness. Attention could be directed  both by symbolic cues in the subject's spared ¢eld  of vision or cues presented in his blind ¢eld. Cues  in his blind ¢eld were even e¡ective in directing  his attention to a second location remote from that  at which the cue was presented. These indirect cues  were e¡ective whether or not they themselves  elicited non-visual awareness. We concluded that the  spatial selection of information by an attentional  mechanism and its entry into conscious experience  cannot be one and the same process.}  }  @Article{klein.EA:00,  author = {Isabelle Klein and Anne-Lise Paradis and  Jean-Baptiste Poline and Stephen M. Kosslyn},  title = {Transient Activity in the Human Calcarine Cortex  During Visual-Mental Imagery: An Event-Related fMRI  Study},  journal = JoCN,  year = 2000,  volume = 12,  number = {Suppl.2},  pages = {15--23},  abstract = {Although it is largely accepted that visual-mental  imagery and perception draw on many of the same  neural structures, the existence and nature of  neural processing in the primary visual cortex (or  area V1) during visual imagery remains  controversial. We tested two general hypotheses:  The first was that V1 is activated only when images  with many details are formed and used, and the  second was that V1 is activated whenever images are  formed, even if they are not necessarily used to  perform a task. We used event-related functional  magnetic resonance imaging (ER-fMRI) to detect and  char- acterize the activity in the calcarine sulcus  (which contains the primary visual cortex) during  single instances of mental imagery. The results  revealed reproducible transient activity in this  area whenever participants generated or evaluated a  mental image. This transient activity was strongly  enhanced when participants evaluated characteristics  of objects, whether or not details actually needed  to be extracted from the image to perform the task.  These results show that visual imagery processing  commonly involves the earliest stages of the visual  system}  }  @Article{kleiser.EA:01,  author = {R. Kleiser and J. Wittsack and M. Niedeggen and  R. Goebel and P. Stoerig},  title = {Is {V1} Necessary for Conscious Vision in Areas of  Relative Cortical Blindness?},  journal = NI,  year = 2001,  volume = 13,  number = 4,  pages = {654--661},  abstract = {Visual field defects result from postgeniculate  lesions. It is generally assumed that absolute  defects are caused by total destruction or  denervation of primary visual cortex (V1) and that  the degraded but conscious vision that remains or  returns in relative or partial defects is mediated  by compromised V1 cortex that retains a sufficiently  large population of functional neurons. We here  report the results of three patients with  long-standing postgeniculate lesions who underwent  functional magnetic resonance imaging while their  partial defect was stimulated with high-contrast  reversing checkerboard stimuli. Although the  stimulation evoked conscious visual impressions in  all three, in only one patient did it activate  perilesional V1. In the other two we found no  evidence for perilesional activation, indicating  that some conscious vision may return in the absence  of functional ipsilesional V1.}  }  @Article{lamme:03,  author = {Victor A.F. Lamme},  title = {Why visual attention and awareness are different},  journal = TICS,  year = 2003,  volume = 7,  number = 1,  pages = {12--18},  abstract = {Now that the study of consciousness is warmly  embraced by cognitive scientists, much confusion  seems to arise between the concepts of visual  attention and visual awareness. Often, visual  awareness is equated to what is in the focus of  attention. There are, however, two sets of arguments  to separate attention from awareness: a  psychological/theoretical one and a neurobiological  one. By combining these arguments I present  definitions of visual attention and awareness that  clearly distinguish between the two, yet explain why  attention and aware- ness are so intricately  related. In fact, there seems more overlap between  mechanisms of memory and awareness than between  those of attention and awareness.}  }  @Article{lamme:06,  author = {Victor A.F. Lamme},  title = {Towards a true neural stance on consciousness},  journal = TICS,  year = 2006,  volume = 10,  number = 11,  pages = {494--501},  abstract = {Consciousness is traditionally defined in mental or  psychological terms. In trying to find its neural  basis, introspective or behavioral observations are  considered the gold standard, to which neural  measures should be fitted. I argue that this poses  serious problems for under- standing the mind–brain  relationship. To solve these problems, neural and  behavioral measures should be put on an equal  footing. I illustrate this by an example from visual  neuroscience, in which both neural and behavioral  arguments converge towards a coherent scientific  definition of visual consciousness. However, to  accept this definition, we need to let go of our  intuitive or psychological notions of conscious  experience and let the neuroscience arguments have  their way. Only by moving our notion of mind towards  that of brain can progress be made.}  }  @Article{lamme.roelfsema:00,  author = {Victor A.F. Lamme and Pieter R. Roelfsema},  title = {The distinct modes of vision offered by feedforward  and recurrent processing},  journal = TIN,  year = 2000,  volume = 23,  pages = {571--579},  abstract = {An analysis of response latencies shows that when an  image is presented to the visual system, neuronal  activity is rapidly routed to a large number of  visual areas. However, the activity of cortical  neurons is not determined by this feedforward sweep  alone. Horizontal connections within areas, and  higher areas providing feedback, result in dynamic  changes in tuning.The differences between  feedforward and recurrent processing could prove  pivotal in understanding the distinctions between  attentive and pre-attentive vision as well as  between conscious and unconscious vision. The  feedforward sweep rapidly groups feature  constellations that are hardwired in the visual  brain, yet is probably incapable of yielding visual  awareness; in many cases, recurrent processing is  necessary before the features of an object are  attentively grouped and the stimulus can enter  consciousness}  }  @Article{laws.EA:05,  author = {Keith R. Laws and Tim M. Gale and Verity C. Leeson  and John R. Crawford},  title = {When is category specific in {Alzheimer's} disease?},  journal = {Cortex},  year = 2005,  volume = 41,  pages = {452--463},  abstract = {Mixed findings have emerged concerning whether  category-specific disorders occur in Alzheimer’s  disease. Factors that may contribute to these  inconsistencies include: ceiling effects / skewed  distributions for control data in some studies;  differences in the severity of cognitive deficit in  patients; and differences in the type of analysis  (in particular, if and how controls are used to  analyse single case data).We examined picture naming  in Alzheimer’s patients and matched elderly healthy  normal controls in three experiments. These  experiments used stimuli that did and did not  produce ceiling effects / skewed data in  controls. In Experiment 1, we examined for category  effects in individual DAT patients using commonly  used analyses for single cases (χ2 and  z-scores). The different techniques produced quite  different outcomes. In Experiment 2a, we used the  same techniques on a different group of patients  with similar outcomes. Finally, in Experiment 2b, we  examined the same patients but (a) used stimuli that  did not produce ceiling effects / skewed  distributions in healthy controls, and (b) used  statistical methods that did not treat the control  sample as a population. We found that ceiling  effects in controls may markedly inflate the  incidence of dissociations in which living things  are differentially impaired and seriously  underestimate dissociations in the opposite  direction. In addition, methods that treat the  control sample as a population led to inflation in  the overall number of dissociations detected. These  findings have implications for the reliability of  category effects previously reported both in  Alzheimer patients and in other pathologies. In  particular, they suggest that the greater proportion  of living than nonliving deficits reported in the  literature may be an artifact of the methods used.}  }  @Article{li.EA:15,  author = {Fali Li and Yin Tian and Yangsong Zhang and Kan Qiu  and Chunyang Tian and Wei Jing and Tiejun Liu and  Yang Xia and Daqing Guo and Dezhong Yao and Peng Xu},  title = {The enhanced information ow from visual cortex to  frontal area facilitates {SSVEP} response: evidence  from model-driven and data-driven causality  analysis},  journal = SR,  year = 2015,  volume = 5,  pages = 14765,  note = {doi:10.1038/srep14765},  abstract = {The neural mechanism of steady-state visual evoked  potentials (SSVEP) is still not clearly  understood. Especially, only certain frequency  stimuli can evoke SSVEP. Our previous network study  reveals that 8 Hz stimulus that can evoke strong  SSVEP response shows the enhanced linkage strength  between frontal and visual cortex. To further probe  the directed information ow between the two cortex  areas for various frequency stimuli, this paper  develops a causality analysis based on the inversion  of double columns model using particle swarm  optimization (PSO) to characterize the directed  information ow between visual and frontal cortices  with the intracranial rat electroencephalograph  (EEG). The estimated model parameters demonstrate  that the 8 Hz stimulus shows the enhanced  directional information ow from visual cortex to  frontal lobe facilitates SSVEP response, which may  account for the strong SSVEP response for 8 Hz  stimulus. Furthermore, the similar nding is  replicated by data-driven causality analysis. The  inversion of neural mass model proposed in this  study may be helpful to provide the new causality  analysis to link the physiological model and the  observed datasets in neuroscience and clinical  researches.}  }  @Article{liu.EA:10,  author = {Zhongming Liu and Cristina Rios and Nanyin Zhang and  Lin Yang and Wei Chen and Bin He},  title = {Linear and nonlinear relationships between visual  stimuli, {EEG and BOLD fMRI} signals},  journal = NI,  year = 2010,  volume = 50,  pages = {1054--1066},  abstract = {In the present study, the cascaded interactions  between stimuli and neural and hemodynamic responses  were modeled using linear systems. These models  provided the theoretical hypotheses that were tested  against the electroencephalography (EEG) and blood  oxygen level dependent (BOLD) functional magnetic  resonance imaging (fMRI) data recorded from human  subjects during prolonged periods of repeated visual  stimuli with a variable setting of the  inter-stimulus interval (ISI) and visual  contrast. Our results suggest that (1) neural  response is nonlinear only when ISIb0.2 s, (2) BOLD  response is nonlinear with an exclusively vascular  origin when 0.25bISIb4.2 s, (3) vascular response  nonlinearity reflects a refractory effect, rather  than a ceiling effect, and (4) there is a strong  linear relationship between the BOLD effect size and  the integrated power of event-related synaptic  current activity, after modeling and taking into  account the vascular refractory effect. These  conclusions offer important insights into the  origins of BOLD nonlinearity and the nature of  neurovascular coupling, and suggest an effective  means to quantitatively interpret the BOLD signal in  terms of neural activity. The validated cross-modal  relationship between fMRI and EEG may provide a  theoretical basis for the integration of these two  modalities.}  }  @Article{macpherson.EA:09,  author = {Helen Macpherson and Andrew Pipingas and Richard  Silberstein},  title = {A steady state visually evoked potential  investigation of memory and ageing},  journal = BC,  year = 2009,  volume = 69,  pages = {571--579},  abstract = {Old age is generally accompanied by a decline in  memory performance. Specifically, neuroimaging and  electrophysiological studies have revealed that  there are age-related changes in the neural  correlates of episodic and working memory. This  study investigated age-associated changes in the  steady state visu- ally evoked potential (SSVEP)  amplitude and latency associated with memory  performance. Participants were 15 older (59–67  years) and 14 younger (20–30 years) adults who  performed an object working memory (OWM) task and a  contextual recognition memory (CRM) task, whilst the  SSVEP was recorded from 64 electrode  sites. Retention of a single object in the low  demand OWM task was characterised by smaller frontal  SSVEP amplitude and latency differences in older  adults than in younger adults, indic- ative of an  age-associated reduction in neural  processes. Recognition of visual images in the more  difficult CRM task was accompanied by larger, more  sustained SSVEP amplitude and latency decreases over  tem- poral parietal regions in older adults. In  contrast, the more transient, frontally mediated  pattern of activ- ity demonstrated by younger adults  suggests that younger and older adults utilize  different neural resources to perform recognition  judgements. The results provide support for  compensatory processes in the aging brain; at lower  task demands, older adults demonstrate reduced  neural activity, whereas at greater task demands  neural activity is increased.}  }  @Article{malinowski.EA:07,  author = {Peter Malinowski and Sandra Fuchs and Matthias  M. M{\"u}ller },  title = {Sustained division of spatial attention to multiple  locations within one hemifield},  journal = NeuroLett,  year = 2007,  volume = 414,  pages = {65--70},  abstract = {Attending to a location in space significantly  improves stimulus perception at that  location. Everyday experience requires the  deployment of attention to multiple objects at  different locations. Recent empirical evidence  suggests that the “beam” of attention can be divided  between non- contiguous areas of the visual  field. Whether this is only possible when stimuli  are presented in different hemifields and harder, if  not impossible, when stimuli are in the same  hemifield is an ongoing debate. Here we use an  electrophysiological measure of sustained  attentional resource allocation (the steady-state  visual evoked potential, SSVEP) to address this  question. In combination with behavioural data we  demonstrate that splitting the attentional “beam” is  in principle possible within one hemifield. However,  results showed that task performance was in general  lower for same-hemifield presentation as opposed to  our previous study with different-hemifield  presentation [M.M. Mu ̈ller, P. Malinowski,  T. Gruber, S.A. Hillyard, Sustained division of the  attentional spotlight, Nature 424 (2003)  309–312]. SSVEP amplitude showed a mixed pattern of  results for stimuli presented in the upper versus  lower quadrant of the left visual hemifield under  conditions of attending to two separated  locations. Results are discussed in the light of the  bilateral distribution advantage hypothesis and  differences in stimulus salience between the upper  and lower visual field.}  }  @Article{manoach.EA:03,  author = {Dara S. Manoach},  title = {Prefrontal cortex dysfunction during working memory  performance in schizophrenia: reconciling discrepant  findings},  journal = SCZres,  year = 2003,  volume = 60,  number = {2--3},  pages = {285--298},  abstract = {Working memory (WM) deficits are a persistent,  disabling and relatively treatment-resistant feature  of schizophrenia that may underlie many cognitive  deficits and symptoms. They are associated with  prefrontal cortex dysfunction. While most  neuroimaging studies of WM demonstrate “task-related  hypofrontality” in schizophrenic relative to healthy  subjects, several recent studies have reported equal  or increased prefrontal activity. These findings  challenge central assumptions regarding cognitive  deficits and prefrontal cortex dysfunction in  schizophrenia. The goal of this review is to  reconcile these seemingly discrepant  findings. Methodological factors addressed include  the use of intersubject averaging, WM task  parameters and the reliability of the  measures. Factors intrinsic to schizophrenia and  their relevance to the selection of experimental  methods and the interpretation of group data are  also discussed. Both hypo- and hyperfrontality are  hypothesized to be valid and informative reflections  of prefrontal cortex dysfunction in  schizophrenia. Due to the heterogeneity and  variability of both performance and regional  recruitment in schizophrenia, whether individual  data is considered, the level and type of WM demands  and the composition of the sample with regard to  performance deficits all influence study outcome and  contribute to discrepancies. Although the prefrontal  cortex is consistently implicated in WM deficits,  the basis of its dysfunction and its exact  contribution remain unclear. Future work might focus  on delineating the exact WM processes, domains and  components that are deficient. In addition,  variability in behavior and activation might best be  regarded as intrinsic to schizophrenia and having a  neural basis that requires explanation. In  combination with other techniques, neuroimaging can  identify the neural circuitry responsible for WM  deficits and elucidate the contribution of each  anatomical component.}  }  @Article{marshall.EA:08,  author = {Randolph S. Marshall and John J. Ferrera and Anna  Barnes and Xian Zhang and Katherine A. O'Brien and  Mohamad Chmayssani and Joy Hirsch and Ronald  M. Lazar},  title = {Brain Activity Associated With Stimulation Therapy  of the Visual Borderzone in Hemianopic Stroke  Patients},  journal = NRNR,  year = 2008,  volume = 22,  number = 2,  pages = {136--144},  abstract = {BACKGROUND AND OBJECTIVE: Visual restoration therapy  is a home-based treatment program intended to expand  visual fields of hemianopic patients through  repetitive stimulation of the borderzone adjacent to  the blind field. We hypothesized that the training  itself would induce visual field location-specific  changes in the brain's response to stimuli, a  phenomenon demonstrated in animal experiments but  never in humans with brain injury. METHODS: Six  chronic right hemianopic patients underwent  functional magnetic resonance imaging  (fMRI)--responding to stimuli in the trained visual  borderzone versus the nontrained seeing field before  and after 1 month of visual restoration  therapy. Spatially normalized fMRI time-series data  were analyzed in a fixed-effects group analysis  comparing blood oxygen level dependent (BOLD)  activity in the borderzone versus seeing location at  baseline and at 1 month. Percent BOLD change was  measured to determine each condition's contribution  to the time-by-condition interaction. RESULTS:  There was a significant time by condition  interaction manifested as increased BOLD activity  for borderzone detection relative to seeing  detection after the first month of therapy, which  correlated with a relative improvement in response  times in the borderzone location out-of-scanner. The  right inferior and lateral temporal, right  dorsolateral frontal, bilateral anterior cingulate,  and bilateral basal ganglia showed the greatest  response. CONCLUSION: Visual restoration therapy  appears to induce an alteration in brain activity  associated with a shift of attention from the  nontrained seeing field to the trained  borderzone. The effect appears to be mediated by the  anterior cingulate and dorsolateral frontal cortex  in conjunction with other higher order visual areas  in the occipitotemporal and middle temporal  regions. Demonstration of a visual field-specific  training effect on brain activity provides an  important starting point for understanding the  potential for visual therapy in hemianopia.}  }  @Article{marcel:98,  author = {Anthony J. Marcel},  title = {Blindsight and shape perception: deficit of visual  consciousness or of visual function?},  journal = {Brain},  year = 1998,  volume = 121,  pages = {1565--1588},  abstract = {Two people with homonymous right hemianopias were  tested on a number of measures of non-conscious and  conscious perception of shape in the blind field.  Experiment 1 examined preparatory manual adjustments  in grasping objects. Both subjects performed well  above chance not only in three-dimensional location  but also in preforming the hand to the shape, size  and orientation of objects. In Experiment 2 single  upper-case letters were briefly exposed in the blind  field, and subjects made forced choices among 6  alternatives in the sighted field. Performance  improved over blocks of trials and was above chance,  but not dramatically. In Experiment 3 single  upper-case words were briefly presented in the blind  field, and subjects chose which of two words exposed  after in the intact field was semantically closer.  In Experiment 4 subjects had to give the meaning of  single ambiguous words (e.g. BANK) presented both  visually in the intact field and auditorily. Each  ambiguous word was preceded by a single upper-case  word briefly presented in the blind field, biasing  each meaning on different blocks of trials (e.g.  MONEY and RIVER). In Experiment 3, although results  were in the appropriate direction, they were not  consistently well above chance. By contrast, in  Keywords : blindsight, consciousness, shape  perception Introduction Blindsight is the name given  to tacit residual visual capacities in a scotoma,  such as discrimination and localization of visual  stimuli that neurologial patients deny seeing  (Sanders et al ., 1974). It is often taken by  psychologists and philosophers as a deficit of  consciousness (Natsoulas, 1982; Dennett, 1991). It  has also been traditionally studied as a deficit of  visual function, especially with regard to the two-  visual-systems hypothesis (Schneider, 1967; Po ̈ ppel  et al. , 1973), and data do show several functional  deficits, e.g. in form and colour perception (for  reviews see Weiskrantz, 1986, 1990). This raises  several issues. (i) Is blindsight both a deficit of  visual consciousness and (separately) of visual  function(s), or basically a deficit of just one  kind? (ii) Are © Oxford University Press 1998  Experiment 4 both subjects were consistently  semantically biased to a high degree by words in the  blind field. Experiments 2, 3 and 4 taken together  suggest that indirect techniques (priming) are more  sensitive to showing effects of non-conscious  perception than direct ones (forced- choice). More  importantly the experiments indicate that not only  orientation but curvature, structural descriptions  of component strokes and spatial ordering of letters  are registered non-consciously in the blind  field. Experiment 5 examined after-images in the  blind and sighted fields, showing veridical  conscious perception of shape in the blind field  provided it was accompanied by a shape in the  sighted field which together formed a good Gestalt.  Experiment 6 showed conscious perception of illusory  contours spanning the hemifields induced by Kanizsa  figures. The experiments suggest that aspects of  shape are much better perceived in blindsight than  previously thought, that this is independent of  their use in motor control, that the main deficit in  blindsight is one of consciousness, and that the  loss of conscious vision in the blind field is far  from total. The effects and their relationship to  those in other neuropsychological deficits suggest  an intimate link between perceptual consciousness,  attention and object perception.}  }  @Article{marzi.EA:91,  author = {C.A. Marzi and P. Bisiacchi and R. Nicoletti},  title = {Is interhemispheric transfer of visuomotor  information asymmetric? Evidence from a  meta-analysis},  journal = {Neuropsychologia},  year = 1991,  volume = 29,  number = 12,  pages = {1163--1177},  abstract = {Using a meta-analytic procedure we have analysed 16  studies employing a simple unimanual reaction time  (RT) paradigm and lateralized visual stimuli to  provide an estimate of interhemispheric transfer  time in normal right-handed subjects. We found a  significant overall RT advantage of the left visual  field over the right and of the right hand over the  left. These asymmetries can be explained by a  superiority of the right hemisphere for the  detection of simple visual stimuli and by a  corresponding superiority of the left hemisphere for  the execution of the manual response,  respectively. Alternatively, they may be interpreted  as related to an asymmetry of interhemispheric  transmission of visuomotor information, with  transfer from the right hemisphere (side of stimulus  entry) to the left (side of response generation)  faster than in the reverse direction. Although a  direct test of these hypotheses is still lacking, we  think that the evidence available is more in keeping  with the latter possibility.}  }  @Article{marzi.EA:06,  author = {C.A. Marzi and F. Mancini and T. Metitieri and  S. Savazzi},  title = {Retinal eccentricity effects on reaction time to  imagined stimuli},  journal = {Neuropsychologia},  year = 2006,  volume = 44,  number = 8,  pages = {1489--1495},  abstract = {To cast light on the possible neural substrate of  visual imagery we tested normal participants and one  hemianopic patient on simple reaction time (RT) to  real and imagined visual stimuli. In one experiment  participants were to detect as quickly as possible a  luminous square presented at one out of two  different retinal eccentricities. A well known  effect with visual stimuli is that RT is slower for  peripheral versus central stimuli. We found that  imagined stimuli showed an eccentricity effect  similar to that obtained with real stimuli. However,  this was not the case in a patient with a hemianopic  visual field loss (quadrantanopia) as a result of  damage to the optic radiation. Even though the  patient showed no difficulty in imaging stimuli in  the affected hemifield she did not show an  eccentricity effect as was the case in her intact  side. In a second experiment, normal participants  showed faster RT to stimuli of larger size with  either real or imagined stimuli. Overall, these  results show that visual perception and imagination  share a similar visuotopic organisation that is  disrupted following deafferentation of the visual  cortex.}  }  @Article{maus.EA:10,  author = {B{\"a}rbel Maus and Gerard J.P. {Van Breukelen} and  Rainer Goebel and Martijn P.F. Berger},  title = {Optimization of blocked designs in {fMRI} studies},  journal = {Psychometrika},  year = 2010,  volume = 75,  number = 2,  pages = {373--390},  abstract = {Blocked designs in functional magnetic resonance  imaging (fMRI) are useful to localize functional  brain areas. A blocked design consists of different  blocks of trials of the same stimulus type and is  charac- terized by three factors: the length of  blocks, i.e., number of trials per blocks, the  ordering of task and rest blocks, and the time  between trials within one block. Optimal design  theory was applied to find the opti- mal combination  of these three design factors. Furthermore,  different error structures were used within a  general linear model for the analysis of fMRI data,  and the maximin criterion was applied to find  designs which are robust against misspecification of  model parameters.}  }  @Article{mazzi.EA:14,  author = {Chiara Mazzi and Francesca Mancini and Silvia  Savazzi},  title = {Can {IPS} reach visual awareness without {V1}?  Evidence from {TMS} in healthy subjects and  hemianopic patients},  journal = {Neuropsychologia},  year = 2014,  volume = 64,  pages = {134--144},  abstract = {TMS of the occipital cortex can elicit conscious  visual percepts, the so-called phosphenes, i.e. the  experience of flashes of light in the absence of an  external stimulus. Previous evidence has shown that  phosphenes can be elicited by TMS of the  intraparietal sulcus (IPS). Here, we tested whether  IPS-phosphenes are generated without any  contribution from the occipital cortex. In  Experiment 1, healthy subjects showed that parietal  phosphenes were reliable different in terms of  threshold, highly variable in size and, in contrast  to the occipital ones, they did not change in  lateralization within the visual field. Moreover, to  strengthen these results and to ascertain whether  feedback to primary visual cortex is necessary for  visual awareness to emerge, in Experiment 2, we  induced phosphenes by applying TMS over the parietal  cortex of the lesioned hemisphere in two hemianopic  patients with a complete destruction of primary  visual cortex. For both patients we found that they  could perceive phosphenes, that a psychophysical  threshold function could be created and that this  function was not different from that obtained with  healthy subjects. In addition, they could reliably  draw the phosphenes and position them within their  blind field and they could reliably score the  phenomenical characteristics of their phosphenes  both in terms of vividness and brightness in a  manner similar to healthy subjects. The two  brain-damaged patients have thus shown that  stimulation of visual areas within the intraparietal  sulcus of their lesioned hemisphere, can, in fact,  elicit conscious visual percepts. These results seem  to demonstrate that TMS-induced visual phenomenology  over IPS can generate conscious visual percepts and,  more importantly, that it does so independently of  any contributions from the primary visual cortex.}  }  @Article{mcgonigle.EA:00,  author = {D. J. {McGonigle} and A. M. Howseman and  B. S. Athwal and K. J. Friston and  R. S. J. Frackowiak and A. P. Holmes},  title = { Variability in {fMRI}: An Examination of  Intersession Differences},  journal = NI,  year = 2000,  volume = 11,  pages = {708--734},  abstract = {The results from a single functional magnetic reso-  nance imaging session are typically reported as  indic- ative of the subject’s functional  neuroanatomy. Under- lying this interpretation is  the implicit assumption that there are no responses  specific to that particular session, i.e., that the  potential variability of response between sessions  is negligible. The present study sought to examine  this assumption empirically. A total of 99 sessions,  comprising 33 repeats of simple motor, visual, and  cognitive paradigms, were collected over a period of  2 months on a single male subject. For each  paradigm, the inclusion of session-by-condition  inter- actions explained a significant amount of  error vari- ance (P < 0.05 corrected for multiple  comparisons) over a model assuming a common  activation magni- tude across all sessions. However,  many of those vox- els displaying significant  session-by-condition interac- tions were not seen in  a multisession fixed-effects analysis of the same  data set; i.e., they were not acti- vated on average  across all sessions. Most voxels that were both  significantly variable and activated on av- erage  across all sessions did not survive a random-  effects analysis (modeling between-session  variance). We interpret our results as demonstrating  that correct inference about subject responses to  activation tasks can be derived through the use of a  statistical model which accounts for both within-  and between-session variance, combined with an  appropriately large ses- sion sample size. If  researchers have access to only a single session  from a single subject, erroneous conclu- sions are a  possibility, in that responses specific to this  single session may be claimed to be typical  responses for this subject.}  }  @Article{mcgonigle:12,  author = {David J. {McGonigle}},  title = {Test–retest reliability in {fMRI}: Or how I learned  to stop worrying and love the variability},  journal = NI,  year = 2012,  volume = 62,  pages = {1116--1120},  abstract = {This paper briefly recounts the details of one of  the first studies of whole brain, single subject  variability in functional Magnetic Resonance Imaging  (fMRI). In this paper I will not attempt to provide  an over- view of fMRI reliability, but will instead  revisit the choices we made in performing our  experiment as we did.}  }  @Article{megumi.EA:15,  author = {Fukuda Megumi and Bahador Bahrami and Ryota Kanai  and Geraint Rees},  title = {Brain activity dynamics in human parietal regions  during spontaneous switches in bistable perception},  journal = NI,  year = 2015,  volume = 107,  pages = {190--197},  abstract = {The neural mechanisms underlying conscious visual  perception have been extensively investigated using  bistable perception paradigms. Previous functional  magnetic resonance imaging (fMRI) and transcranial  magnetic stimula- tion (TMS) studies suggest that  the right anterior superior parietal (r-aSPL) and  the right posterior superior parietal lobule  (r-pSPL) have opposite roles in triggering  perceptual reversals. It has been proposed that  these two areas are part of a hierarchical network  whose dynamics determine perceptual  switches. However, how these two parietal regions  interact with each other and with the rest of the  brain during bistable perception is not known. Here,  we investigated such a model by recording brain  activity using fMRI while participants viewed a  bistable structure- from-motion stimulus. Using  dynamic causal modeling (DCM), we found that  resolving such perceptual ambiguity was specifically  associated with reciprocal interactions between  these parietal regions and V5/MT. Strikingly, the  strength of bottom-up coupling between V5/MT to  r-pSPL and from r-pSPL to r-aSPL predicted  individual mean dominance duration. Our findings are  consistent with a hierarchical predictive coding  model of parietal involve- ment in bistable  perception and suggest that visual information  processing underlying spontaneous perceptual  switches can be described as changes in connectivity  strength between parietal and visual cortical  regions.}  }  @Article{mele.EA:08,  author = {Sonia Mele and Silvia Savazzi and Carlo A. Marzi and  Giovanni Berlucchi},  title = {Reaction time inhibition from subliminal cues: Is it  related to inhibition of return?},  journal = {Neuropsychologia},  year = 2008,  volume = 46,  pages = {810--819},  abstract = {Task-irrelevant visual cues with near zero  visibility proved apt to retard reaction time for  the detection of supraliminal visual targets  presented at the cued location. The time course of  the effect was similar to that of the so-called  inhibition-of return (IOR), which is assumed to be  due to the withdrawal of attention from the  inhibited location. However the present subliminal  cues consistently failed to induce an RT  facilitation prior to the RT inhibition, contrary to  what would be expected if the cue were able to  attract attention to the cued location. Since the RT  inhibition from subliminal cues could not be  attributed to the withdrawal of attention from the  cued location, it can be argued that such cues acted  both outside of consciousness and without the  influence of attention. Therefore, the RT inhibitory  effect seems best accounted for by an automatic,  unconscious and attention-independent  self-inhibition of response tendencies instructed by  irrelevant information, akin to that postulated by  (Eimer, M., & Schlaghecken, F. (1998). Effects of  masked stimuli on motor activation: behavioural and  electrophysiological evidence.}  }  @Article{miki.EA:05,  author = {Atsushi Miki and Grant T. Liu and Edward  J. Modestino and Gabrielle R. Bonhomme and  Chia-Shang J. Liu and John C. Haselgrove},  title = {Decreased Lateral Geniculate Nucleus Activation in  Retrogeniculate Hemianopia Demonstrated by  Functional Magnetic Resonance Imaging at 4Tesla},  journal = {Ophthalmologica},  year = 2005,  volume = 219,  pages = {11--19},  abstract = {Functional magnetic resonance imaging (fMRI) can de-  tect lateral geniculate nucleus (LGN) activation. We  stud- ied LGN function in 5 patients with  retrogeniculate hom- onymous hemianopia using fMRI  at 4.0 Tesla during binocular visual  stimulation. Decreased activation of vi- sual cortex  and LGN on the side of the lesion was ob- served in  all 5 patients.These findings suggest that ret-  rogeniculate lesions are associated with decreased  activation of the LGN, due to retrograde  degeneration or a functional decrease caused by  decreased feedback from ipsilateral visual cortex.}  }  @Book{milner.goodale:95,  author = {A. David Milner and Melvyn A. Goodale},  title = {The visual brain in action},  publisher = {Oxford Science Publications},  year = 1995,  series = {Oxford Psychology Series no.27},  address = {Oxford}}  @Article{morland.EA:02,  author = {A. B. Morland and M. B. Hoffmann and M. Neveu and  G. E. Holder},  title = {Abnormal visual projection in a human albino studied  with functional magnetic resonance imaging and  visual evoked potentials},  journal = JNNP,  year = 2002,  volume = 72,  pages = {523--526},  abstract = {The albino visual pathway is abnormal in that many  fibres from the temporal retina project to the  contralateral visual cortex. The visual projections  in a human albino and a control have been  investigated with fMRI and VEP during independent  visual stimulation of both hemifields. Activity in  the occipital cortex in the normal was contralateral  to the stimulated visual field, whereas it was  contralateral to the stimulated eye in the albino,  independent of the stimulated visual field. Thus,  the albino visual cortex is activated not only by  stimulation in the contralateral visual field, but  also by abnormal input representing the ipsilateral  visual field. These novel findings help elucidate  the nature of albino misrouting. }  }  @Article{morland.EA:04,  author = {Antony B. Morland and Sandra L{\^e} and Erin Carroll  and Michael B. Hoffmann and Alidz Pambakian},  title = {The role of spared calcarine cortex and lateral  occipital cortex in the responses of human  hemianopes to visual motion},  journal = JoCN,  year = 2004,  volume = 16,  number = 2,  pages = {204--218},  abstract = {Some patients, who are rendered perimetrically blind  in one hemifield by cortical lesions, nevertheless  exhibit residual visual capacities within their  field defects. The neural mechanism that mediates  the residual visual responses has remained the topic  of considerable debate. One explanation posits the  subcortical visual pathways that bypass the primary  visual cortex and innervate the extrastriate visual  areas as the substrate that underlies the residual  vision. The other explanation is that small islands  of the primary visual cortex remain intact and  provide the signals for residual vision. We have  performed behavioral and functional magnetic  resonance imaging experiments to investigate the  validity of the two explanations of residual  vision. Our behavioral experiments indicated that of  the seven hemianopes tested, two had the ability to  discriminate the direction of a drifting  grating. This residual visual response was shown  with fMRI to be the result of spared islands of  calcarine cortical activity in one of the  hemianopes, whereas only lateral occipital activity  was documented in the other patient. These results  indicate that the underlying neural correlates of  residual vision can vary between patients. Moreover,  our study emphasizes the necessity of ruling out the  presence of islands of preserved function and  primary visual cortex before assigning residual  visual capacities to the properties of visual  pathways that bypass the primary visual cortex.}  }  @Article{mueller.EA:16,  author = {Matthias M. M{\"u}ller and Mireille Trautmann and  Christian Keitel},  title = {Early Visual Cortex Dynamics during Top–Down  Modulated Shifts of Feature-Selective Attention},  journal = JoCN,  year = 2016,  volume = 28,  number = 4,  pages = {643--655},  annote = {Shifting attention from one color to another color  or from color to another feature dimension such as  shape or orientation is imperative when searching  for a certain object in a cluttered scene. Most  attention models that emphasize feature-based  selection implicitly assume that all shifts in  feature-selective attention underlie identical  temporal dynamics. Here, we re- corded time courses  of behavioral data and steady-state visual evoked  potentials (SSVEPs), an objective  electrophysiological measure of neural dynamics in  early visual cortex to investigate temporal dynamics  when participants shifted attention from color or  orientation toward color or orientation,  respectively. SSVEPs were elicited by four random  dot kinematograms that flickered at different  frequencies. Each random dot kinematogram was  composed of dashes that uniquely combined two fea-  tures from the dimensions color (red or blue) and  orientation (slash or backslash). Participants were  cued to attend to one feature (such as color or  orientation) and respond to coherent motion targets  of the to-be-attended feature. We found that shifts  toward color occurred earlier after the shifting cue  com- pared with shifts toward orientation,  regardless of the original feature (i.e., color or  orientation). This was paralleled in SSVEP amplitude  modulations as well as in the time course of behav-  ioral data. Overall, our results suggest different  neural dynamics during shifts of attention from  color and orientation and the respective shifting  destinations, namely, either toward color or toward  orientation.}  }  @Article{nelles.EA:02,  author = {Gereon Nelles and Guido Widman and Armin {de Greiff}  and Anette Meistrowitz and Albena Dimitrova and  Johannes Weber and Michael Forsting and Joachim  Esser and H. Christoph Diener},  title = {Brain Representation of Hemifield Stimulation in  Poststroke Visual Field Defects},  journal = {Stroke},  year = 2002,  volume = 33,  pages = {1286--1293},  abstract = {Background and Purpose— Plasticity in extended,  parallel, or reciprocal operating networks is well  recognized. Changes in neuronal activity after  lesions to distinct localized structures, such as  the primary visual cortex, are less well  characterized. We investigated the cortical  reorganization in patients with poststroke visual  field defects using blood oxygen level–dependent  functional MRI. Methods— Brain activation was  measured in 7 patients with a single occipital  cortical lesion and partially recovered hemianopia  and in 7 age-matched control subjects. Differences  in activation between rest and visual hemifield  stimulation were assessed with statistical  parametric mapping (SPM’99). Results— In normal  subjects, significant activation was found in the  contralateral primary visual cortex and bilaterally  in the extrastriate cortex. During hemifield  stimulation of the unaffected side of stroke  patients, a similar pattern was found compared with  that seen in control subjects. During stimulation of  the hemianopic side, bilateral activation was seen  within the extrastriate cortex, stronger in the  ipsilateral hemisphere. The primary visual cortex  was not significantly activated in either hemisphere  during stimulation of the hemianopic side.  Conclusions— Visual field defects after stroke are  associated with bilateral activation of the  extrastriate visual cortex. This pattern of  activation indicates altered neuronal activity in  the visual system. Further investigation is  necessary to determine the relationship between  functional reorganization and recovery of lost  visual function after poststroke hemianopia. }  }  @Article{nelles.EA:07,  author = {Gereon Nelles and Armin {de Greiff} and Anja  Pscherer and Michael Forsting and Horst Gerhard and  Joachim Esser and H. Christoph Diener},  title = {Cortical activation in hemianopia after stroke},  journal = NeuroLett,  year = 2007,  volume = 426,  number = 1,  pages = {34--38},  abstract = {Changes in neuronal activity of the visual cortex  have been described in patients with hemianopia. The  anatomical areas that are involved in neuroplastic  changes have not been studied in a larger group of  stroke patients with a homogenous structural  pathology of the visual cortex. Brain activation was  measured in 13 patients with a single ischemic  lesion of the striate cortex and partially recovered  hemianopia and in 13 age-matched control subjects  using blood oxygen level dependent (BOLD) functional  magnetic resonance imaging (fMRI). Differences in  activation between rest and visual hemifield  stimulation were assessed with statistical  parametric mapping using group and multi-group  studies. In normal subjects, the most significant  activation was found in the contralateral primary  visual cortex (area 17) and bilaterally in the  extrastriate cortex (areas 18 and 19). In patients,  these areas were also activated when the intact  hemifield was stimulated. During stimulation of the  hemianopic side, bilateral activation was seen  within the extrastriate cortex, stronger in the  ipsilateral (contralesional) hemisphere. Stimulation  of the hemianopic visual field is associated with  ipsilateral activation of the extrastriate visual  cortex. This pattern of activation suggests  extensive neuronal plasticity within the visual  cortex after postgeniculate ischemic lesions and may  have implications for therapeutic interventions.}  }  @Article{norcia.EA:15,  author = {Anthony M. Norcia and L. Gregory Appelbaum and  Justin M. Ales and Benoit R. Cottereau and Bruno  Rossion},  title = {The steady-state visual evoked potential in vision  research: A review},  journal = JoVis,  year = 2015,  volume = 15,  number = 6,  pages = {1--46},  abstract = {Periodic visual stimulation and analysis of the  resulting steady-state visual evoked potentials were  first introduced over 80 years ago as a means to  study visual sensation and perception. From the  first single-channel recording of responses to  modulated light to the present use of sophisticated  digital displays composed of complex visual stimuli  and high-density recording arrays, steady-state  methods have been applied in a broad range of  scientific and applied settings. The purpose of this  article is to describe the fundamental stimulation  paradigms for steady-state visual evoked potentials  and to illustrate these principles through research  findings across a range of applications in vision  science.}  }  @Article{,  author = {Robert Oostenveld and Pascal Fries and Eric Maris  and Jan-Mathijs Schoffelen},  title = {{FieldTrip}: Open Source Software for Advanced  Analysis of {MEG, EEG}, and Invasive  Electrophysiological Data},  journal = CIN,  year = {2011},  OPTkey = {},  OPTvolume = {},  OPTnumber = {},  OPTpages = {},  OPTmonth = {},  OPTnote = {doi:10.1155/2011/156869},  abstract = {This paper describes FieldTrip, an open source  software package that we developed for the analysis  of MEG, EEG, and other electrophysiological  data. The software is implemented as a MATLAB  toolbox and includes a complete set of consistent  and user- friendly high-level functions that allow  experimental neuroscientists to analyze experimental  data. It includes algorithms for simple and advanced  analysis, such as time-frequency analysis using  multitapers, source reconstruction using dipoles,  distributed sources and beamformers, connectivity  analysis, and nonparametric statistical permutation  tests at the channel and source level. The  implementation as toolbox allows the user to perform  elaborate and structured analyses of large data sets  using the MATLAB command line and batch  scripting. Furthermore, users and developers can  easily extend the functionality and implement new  algorithms. The modular design facilitates the reuse  in other software packages.}  }  @Article{palomares.EA:12,  author = {Melanie Palomares and Justin M. Ales and Alex  R. Wade and Benoit R. Cottereau and Anthony  M. Norcia},  title = {Distinct effects of attention on the neural  responses to form and motion processing: A {SSVEP}  source-imaging study},  journal = JoVis,  year = 2012,  volume = 12,  number = 10,  pages = {1--14},  abstract = {We measured neural responses to local and global  aspects of form and motion stimuli using  frequency-tagged, steady-state visual evoked  potentials (SSVEPs) combined with magnetic resonance  imaging (MRI) data. Random dot stimuli were used to  portray either dynamic Glass patterns (Glass, 1969)  or coherent motion displays. SSVEPs were used to  estimate neural activity in a set of fMRI-defined  visual areas in each subject. To compare activity  associated with local versus global processing, we  analyzed two frequency components of the SSVEP in  each visual area: the high temporal frequency at  which the local dots were updated (30 Hz) and the  much lower frequency corresponding to updates in the  global structure (0.83 Hz). Local and global  responses were evaluated in the context of two  different behavioral tasks—subjects had to either  direct their attention toward or away from the  global coherence of the stimuli. The data show that  the effect of attention on global and local  responses is both stimulus and visual area  dependent. When attention was directed away from  stimulus coherence, both local and global responses  were higher in the coherent motion than Glass  pattern condition. Directing attention to coherence  in Glass patterns enhanced global activity in areas  LOC, hMTþ, V4, V3a, and V1, while attention to  global motion modulated responses by a smaller  amount in a smaller set of areas: V4, hMTþ, and  LOC. In contrast, directing attention towards  stimulus coherence weakly increased local responses  to both coherent motion and Glass patterns. These  results suggest that visual attention differentially  modulates the activity of early visual areas at both  local and global levels of structural encoding.}  }  @Article{paramei.sabel:08,  author = {Galina V. Paramei and Bernhard A. Sabel},  title = {Contour-integration deficits on the intact side of  the visual field in hemianopia patients},  journal = BBR,  year = 2008,  volume = 188,  pages = {109--124},  abstract = {Objective: Visual impairments in hemianopia are  thought to be exclusively caused by the reduced  visual field size. However, the primary lesion may  affect the contralateral hemisphere through damage  of interhemispheric projections. The question  therefore arises if the presumed “intact” hemifield  is perceptually impaired. Methods: Three hemianopia  patients and three matched controls carried out a  Yes/No figure detection task with their intact side  of the visual field. The figure (square) contours  were composed of non-contiguous Gabor patches  embedded in a random patch array of different  background densities (low, Δ = 2; high, Δ =  1). Response accuracy and reaction times were  recorded. Results: A temporal–parietal patient  revealed figure detection impairments, with accuracy  rate,77\% (Δ=2) and 53\% (Δ=1),belowcompared control  values. An occipital patient was comparable to his  match: 99\% (Δ = 2); 84\% (Δ = 1). Both patients  exhibited frequent false alarms to random patterns  and required longer presentation times to perform  the task. In the third patient, with optic tract  lesion, figure detection was nearly normal at low  density (92\%, Δ = 2) but impaired at noisy  background (62\%, Δ = 1). Conclusion: The “intact”  visual field in hemianopes is impaired in detection  of incomplete figures embedded in a noisy  background. This deficit may be caused by damage to  higher visual centers and/or loss of  interhemispheric interactions.}  }  @Article{pastor.EA:03,  author = {Maria A. Pastor and Julio Artieda and Javier Arbizu  and Miguel Valencia and Jose C. Masdeu},  title = {Human Cerebral Activation during Steady-State  Visual-Evoked Responses},  journal = JoN,  year = 2003,  volume = 23,  number = 3,  pages = {11621--11627},  abstract = {Flicker stimuli of variable frequency (2-90 Hz)  elicit a steady-state visual-evoked response (SSVER)  in the electroencephalogram (EEG) with the same  frequency as the stimulus. In humans, the amplitude  of this response peaks at ∼15 Hz, decreasing at  higher stimulation frequencies. It was not known  whether this peak response corresponds to increased  synaptic activity in the visual cortex or to other  mechanisms [for instance, the temporal coherence  (phase summation) of evoked responses]. We studied  the SSVER in 16 normal volunteers by means of visual  stimulation at 14 different frequencies (from 5 to  60 Hz) while recording the EEG. In nine subjects of  the group, we measured regional cerebral blood flow  (rCBF) with positron emission tomography (PET)-H215O  at rest and during visual stimulation at five  different frequencies: 5, 10, 15, 25, and 40 Hz. We  confirmed that the amplitude of the SSVER in  occipital regions peaks at 15 Hz  stimulation. Applying to the PET rCBF data a  contrast weighted by the amplitude of the SSVER, we  determined that the primary visual cortex rCBF  follows an activation pattern similar to the  SSVER. This finding suggests that the amplitude of  the SSVER corresponds to increased synaptic  activity, specifically in Brodmann's area  17. Additionally, this study showed that visual  stimulation at 40 Hz causes selective activation of  the macular region of the visual cortex, and that a  region in the dorsal aspect of the Crus I lobule of  the left cerebellar hemisphere is activated during  repetitive visual stimulation. }  }  @Article{perez.EA:13,  author = {C{\'e}line Perez and Carole Peyrin and C{\'e}line  Cav{\'e}zianand Olivier Coubard and Florent Caetta  and Noa Raz and Netta Levin and Ga{\"e}lle Doucet  and Fr{\'e}d{\'e}ric Andersson and Micha{\"e}l  Obadia and Olivier Gout and Fran{\,c}oise H{\'e}ran  and Julien Savatovsky and Sylvie Chokron},  title = {An {fMRI} Investigation of the Cortical Network  Underlying Detection and Categorization Abilities in  Hemianopic Patients},  journal = BT,  year = 2013,  volume = 26,  number = 2,  pages = {264--277},  note = {DON'T HAVE},  abstract = {The current study aims to investigate visual scene  perception and its neuro-anatomical correlates for  stimuli presented in the central visual field of  patients with homonymous hemianopia, and thereby to  assess the effect of a right or a left occipital  lesion on brain reorganization. Fourteen healthy  participants, three left brain damaged (LBD)  patients with right homonymous hemianopia and five  right brain damaged (RBD) patients with left  homonymous hemianopia performed a visual detection  task (i.e. “Is there an image on the screen?”) and a  categorization task (i.e. “Is it an image of a  highway or a city?”) during a block-designed  functional magnetic resonance imaging recording  session. Cerebral activity analyses of the posterior  areas—the occipital lobe in particular—highlighted  bi-hemispheric activation during the detection task  but more lateralized, left occipital lobe activation  during the categorization task in healthy  participants. Conversely, in patients, the same  network of activity was observed in both  tasks. However, LBD patients showed a predominant  activation in their right hemisphere (occipital lobe  and posterior temporal areas) whereas RBD patients  showed a more bilateral activation (in the occipital  lobes). Overall, our preliminary findings suggest a  specific pattern of cerebral activation depending on  the task instruction in healthy participants and  cerebral reorganization of the posterior areas  following brain injury in hemianopic patients which  could depend upon the side of the occipital lesion.}  }  @Article{perez.chokron:14,  author = {C{\'e}line Perez and Sylvie Chokron},  title = {Rehabilitation of homonymous hemianopia: insight  into blindsight},  journal = FIN,  year = 2014,  volume = 8,  number = 82,  note = {doi:10.3389/fnint.2014.00082},  annote = {Strong evidence of considerable plasticity in  primary sensory areas in the adult cortex, and of  dramatic cross-modal reorganization in visual areas,  after short- or long-term visual deprivation has  recently been reported. In the context of patient  rehabilitation, this scientifically challenging  topic takes on urgent clinical relevance, especially  given the lack of information about the role of such  reorganization on spared or newly emerged visual  performance. Amongst the most common visual field  defects found upon unilateral occipital damage of  the primary visual cortex is homonymous hemianopia  (HH), a perfectly symmetric loss of vision in both  eyes. Traditionally, geniculostriate lesions were  considered to result in complete and permanent  visual loss in the topographically related area of  the visual field (Huber, 1992). However, numerous  studies in monkeys, and later, in humans, have  demonstrated that despite destruction of the striate  cortex, or even following a hemispherectomy, some  patients retain a certain degree of unconscious  visual function, known as blindsight. Accordingly,  there have recently been attempts to restore visual  function in patients by stimulating unconscious  preserved blindsight capacities. Herein we review  different visual rehabilitation techniques designed  for brain-damaged patients with visual field  loss. We discuss the hypothesis that explicit  (conscious) visual detection can be restored in the  blind visual field by harnessing implicit  (unconscious) visual capacities. The results that we  summarize here underline the need for early  diagnosis of cortical visual impairment (CVI), and  the urgency in rehabilitating such deficits, in  these patients. Based on the research precedent, we  explore the link between implicit (unconscious)  vision and conscious perception and discuss possible  mechanisms of adaptation and plasticity in the  visual cortex.}  }  @Article{perlstein.EA:03,  author = {William M. Perlstein and Michael A. Cole and Michael  Larson and Kiesa Kelly and Paul Seignourel and  Andreas Keil},  title = {Steady-state visual evoked potentials reveal  frontally-mediated working memory activity in  humans},  journal = NeuroLett,  year = 2003,  volume = 342,  pages = {191--195},  abstract = {Steady-state visual evoked potentials (SSVEPs)  reflect power changes at the stimulus driving  frequency and have been used to assess brain  activity reflecting cognitive processing. Only one  study has demonstrated SSVEP modulation associated  with working memory (WM), and none have compared the  spatial localization of SSVEP modulations during WM  performance with other brain imaging methods. Here  we examined WM-related activity recorded with  dense-array SSVEPs, analyzed using low resolution  electromagnetic tomography, and compared the results  to our previous findings using functional magnetic  resonance imaging (fMRI). WM was associated with  increased SSVEP activity over the right dorsolateral  prefrontal cortex, paralleling our previous fMRI  findings. Frontal WM-related SSVEP power correlated  selectively with task performance. These results  demonstrate the utility of SSVEPs for studying  representational aspects of cognition.}  }  @Article{pessoa.EA:03,  author = {Luiz Pessoa and Sabine Kastner and Leslie  G. Ungerleider},  title = {Neuroimaging Studies of Attention: From Modulation  of Sensory Processing to Top-Down Control},  journal = JoN,  year = 2003,  volume = 23,  number = 10,  pages = {3990--3998},  abstract = {What we perceive depends critically on where we  direct our attention. For example, attention to a  location dramatically improves the accuracy and  speed of detecting a target at that  location. Attention has been shown not only to  increase perceptual sensitivity for target  discrimination but also to reduce the interference  caused by nearby distracters. Moreover, attention is  highly flexible and can be deployed in a manner that  best serves the organism's momentary behavioral  goals: to locations, to visual features, or to  objects. Attention can also be based on internal  goals (e.g., finding a familiar face in the crowd)  or depend on the external environment (e.g., as when  a loud alarm sounds). A large body of evidence  exists from both neurophysiology and functional  magnetic resonance imaging (fMRI) concerning the  neural mechanisms subserving attention. In this  review, we focus on the parallels and differences  that exist between findings revealed by these two  techniques. We also highlight important  contributions to the field of attention that have  been brought about by fMRI studies. We discuss the  effects of attention on activity within areas of  visual cortex, as well as the evidence for the  existence of source areas exerting top-down  attentional control that are thought to provide  biasing signals that are observed in target areas in  visual cortex. In general, in linking single-cell  physiology and fMRI studies, it is assumed that the  effects observed with fMRI directly reflect the  summed responses of large populations of  neurons. However, recent studies, including the  seminal work by Logothetis and colleagues  (Logothetis et al., 2001; Logothetis, 2003), have  shown that the mapping between these two domains is  not so simple. It has been suggested that blood  oxygenation level-dependent (BOLD) signals reflect  the incoming inputs into an area as well as the  processing of this input information by the local  cortical circuitry, including excitatory and  inhibitory interneurons. Thus, in some cases fMRI  may reveal significant activation that may have no  counterpart in single-cell physiology. In such  cases, local field potentials (LFPs), which also  reflect both input signals and local intracortical  processing, appear to correlate better with fMRI  signals. Interestingly, the acquisition of LFPs by  physiologists is becoming more prevalent (Fries et  al., 2001) and may help bridge the gap between  results from single-cell and fMRI studies. }  }  @Article{peterson.EA:14,  author = {Dwight J. Peterson and Gennadiy Gurariy and  Gabriella G. Dimotsantos and Hector Arciniega and  Marian E. Berryhill and Gideon P. Caplovitz},  title = {The steady-state visual evoked potential reveals  neural correlates of the items encoded into visual  working memory},  journal = {Neuropsychologia},  year = 2014,  volume = 63,  pages = {145--153},  abstract = {Visual working memory (VWM) capacity limitations are  estimated to be ~4 items. Yet, it remains unclear  why certain items from a given memory array may be  successfully retrieved from VWM and others are  lost. Existing measures of the neural correlates of  VWM cannot address this question because they  measure the aggregate processing of the entire  stimulus array rather than neural signatures of  individual items. Moreover, this cumulative  processing is usually measured during the delay  period, thereby reflecting the allocation of neural  resources during VWM maintenance. Here, we use the  steady- state visual evoked potential (SSVEP) to  identify the neural correlates of individual stimuli  at VWM encoding and test two distinct hypotheses:  the focused-resource hypothesis and the  diffuse-resource hypothesis, for how the allocation  of neural resources during VWM encoding may  contribute to VWM capacity limitations. First, we  found that SSVEP amplitudes were larger for stimuli  that were later remembered than for items that were  subsequently forgotten. Second, this pattern  generalized so that the SSVEP amplitudes were also  larger for the unprobed stimuli in correct compared  to incorrect trials. These data are consistent with  the diffuse-resource view in which attentional  resources are broadly allocated across the whole  stimulus array. These results illustrate the  important role encoding mechanisms play in limiting  the capacity of VWM.}  }  @Article{plow.EA:12,  author = {Ela B. Plow and Souzana N. Obretenova and Felipe  Fregni and Alvaro Pascual-Leone and Lotfi  B. Merabet},  title = {Comparison of Visual Field Training for Hemianopia  With Active Versus Sham Transcranial Direct Cortical  Stimulation},  journal = NNR,  year = 2012,  volume = 26,  number = 6,  pages = {616--626},  abstract = {Background. Vision Restoration Therapy (VRT) aims to  improve visual field function by systematically  training regions of residual vision associated with  the activity of suboptimal firing neurons within the  occipital cortex. Transcranial direct current  stimulation (tDCS) has been shown to modulate  cortical excitability. Objective. Assess the  possible efficacy of tDCS combined with  VRT. Methods. The authors conducted a randomized,  double-blind, demonstration-of-concept pilot study  where participants were assigned to either VRT and  tDCS or VRT and sham. The anode was placed over the  occipital pole to target both affected and  unaffected lobes. One hour training sessions were  carried out 3 times per week for 3 months in a  laboratory. Outcome measures included objective and  subjective changes in visual field, recording of  visual fixation performance, and vision-related  activities of daily living (ADLs) and quality of  life (QOL). Results. Although 12 participants were  enrolled, only 8 could be analyzed. The VRT and tDCS  group demonstrated significantly greater expansion  in visual field and improvement on ADLs compared  with the VRT and sham group. Contrary to  expectations, subjective perception of visual field  change was greater in the VRT and sham group. QOL  did not change for either group. The observed  changes in visual field were unrelated to  compensatory eye movements, as shown with fixation  monitoring. Conclusions. The combination of  occipital cortical tDCS with visual field  rehabilitation appears to enhance visual functional  outcomes compared with visual rehabilitation  alone. TDCS may enhance inherent mechanisms of  plasticity associated with training. }  }  @Article{price.EA:06,  author = {Cathy J. Price and Jenny Crinion and Karl  J. Friston},  title = {Design and Analysis of {fMRI} Studies With  Neurologically Impaired Patients},  journal = JMRI,  year = 2006,  volume = 23,  pages = {816--826},  abstract = {Functional neuroimaging can be used to characterize  two types of abnormality in patients with  neurological deficits: abnormal functional  segregation and abnormal functional integration. In  this paper we consider the factors that in- fluence  the experimental design, analysis, and interpreta-  tion of such studies. With respect to experimental  design, we emphasize that: 1) task selection is  constrained to tasks the patient is able to perform  correctly, and 2) the most sensitive designs entail  presenting stimuli of the same type close  together. In terms of data preprocessing, prior to  statistical analysis, we note that structural  pathology may call for constraints on nonlinear  transformations, used by spatial normalization, to  prevent distortion of intact tissue. This means that  one may have to increase spatial smooth- ing to  reduce the impact of inaccurate normalization. Im-  portant issues in statistical modeling concern the  first level of analysis (estimation of activation  within subject), which has to distinguish correct  from incorrect responses. At the second level  (between subjects), inference should be based on  between-subjects variance. Provided that these and  other constraints are met, deficits in functional  segregation are indicated when activation in one or  a set of regions is higher or lower in patients  relative to control subjects. In contrast, deficits  in functional integration are implied when the  influence of one brain region on another is stronger  or weaker in patients relative to control subjects.}  }  @Article{ptak.schnider:11,  author = {Radek Ptak and Armin Schnider},  title = {The attention network of the human brain: Relating  structural damage associated with spatial neglect to  functional imaging correlates of spatial attention},  journal = {Neuropsychologia},  year = 2011,  volume = 49,  pages = {3063--3070},  abstract = {Functional imaging studies of spatial attention  regularly report activation of the intraparietal  sulcus (IPS) and dorsal premotor cortex including  the frontal eye fields (FEF) in tasks requiring  overt or covert shifting of attention. In contrast,  lesion-overlap studies of patients with spatial  neglect – a syndrome character- ized by severe  impairments of spatial attention – show that the  critical damage concerns more ventral regions,  comprising the inferior parietal lobule, the  temporal–parietal junction (TPJ), and the superior  temporal gyrus. We performed voxel-based  lesion-symptom mapping of 29 right-hemisphere stroke  patients, using several performance indices derived  from a cueing task as measures of spatial  attention. In contrast to previous studies, we  focused our analyses on eight regions of interest  defined according to results of previous functional  imaging studies. A direct comparison of neglect with  control patients revealed that neglect was  associated with damage to the TPJ, the middle  frontal gyrus, and the posterior IPS. The latter  region was also a significant predictor of the  degree of contralesional slowing of target detection  and the extent to which ipsilesional distracters  captured attention of neglect patients. Finally,  damage to the FEF and posterior IPS was negatively  correlated with the tendency of neglect patients to  ori- ent attention toward behaviourally relevant  distracters. These findings support the results of  functional imaging studies of spatial attention and  provide evidence for a network account of neglect,  according to which attentional selection of relevant  environmental stimuli and the reorienting of  attention result from dynamic interactions between  the IPS, the dorsal premotor cortex, and the TPJ.}  }  @Article{ptito.EA:99,  author = {M. Ptito and P. Johannsen and J. Faubert and  A. Gjedde},  title = {Activation of Human Extrageniculostriate Pathways  after Damage to Area {V1}},  journal = NI,  year = 1999,  volume = 9,  number = 1,  pages = {97--107},  abstract = {We have used positron emission tomography (PET) to  specify the cortical and subcortical structures  activated following visual stimulation of the  scotomatous field in a patient with an asymmetric  bilateral developmental anomaly of the visual  cortex. Computerized perimetry indicated a left  visual field defect only although MRI and18FDG-PET  scans showed abnormalities in both occipital  lobes. The visual stimuli were semicircular gratings  moving in opposite directions on a dynamic  random-dot background. They were specifically  constructed to eliminate intra- and extraocular  light scatter and to optimize the activation of  extra-striate cortical areas and their projecting  subcortical relays. For anatomical localization PET  images were coregistered to the subject's MRI in  Talairach coordinates. After subtraction of the  baseline conditions from the stimulation conditions,  at-statistic map was created on a voxel-by-voxel  basis. Stimulation of the scotomatous hemifield  yielded significant activations of Brodmann cortical  areas 18–19 and 47 as well as the pulvinar thalami  of the left hemisphere, in addition to a less  prominent activation in the right  hemisphere. Stimulation of the intact hemifield  produced significant activation of Brodmann cortical  areas 30 and 47 of the left hemisphere. These  results suggest that in the absence of area V1,  residual vision observed in the blind hemifield  could be mediated by a retinofugal pathway to  extrastriate cortex via the pulvinar.}  }  @Article{raemaekers.EA:11,  author = {M. Raemaekers and D. P. Bergsmaand R. J. A. {van  Wezel} and G. J. {van der Wildt} and A. V. {van den  Berg}},  title = {Effects of Vision Restoration Training on Early  Visual Cortex in Patients With Cerebral Blindness  Investigated With Functional Magnetic Resonance  Imaging},  journal = JNeurophys,  year = 2011,  volume = 105,  pages = {872--882},  abstract = {Cerebral blindness is a loss of vision as a result  of postchiasmatic damage to the visual  pathways. Parts of the lost visual field can be  restored through training. However, the neuronal  mechanisms through which training effects occur are  still unclear. We therefore assessed  training-induced changes in brain function in eight  patients with cerebral blindness. Visual fields were  measured with perimetry and retinotopic maps were  acquired with functional magnetic resonance imaging  (fMRI) before and after vision restoration  training. We assessed differences in hemodynamic re-  sponses between sessions that represented changes in  amplitudes of neural responses and changes in  receptive field locations and sizes. Perimetry  results showed highly varied visual field recovery  with shifts of the central visual field border  ranging between 1 and 7°. fMRI results showed that,  although retinotopic maps were mostly stable over  sessions, there was a small shift of receptive field  locations toward a higher eccentricity after  training in addition to increases in receptive field  sizes. In patients with bilateral brain activation,  these effects were stronger in the affected than in  the intact hemisphere. Changes in receptive field  size and location could account for limited visual  field recovery ( 1°), although it could not account  for the large increases in visual field size that  were observed in some patients. Furthermore, the  retinotopic maps strongly matched perimetry mea-  surements before training. These results are taken  to indicate that local visual field enlargements are  caused by receptive field changes in early visual  cortex, whereas large-scale improvement cannot be  ex- plained by this mechanism.}  }  @Article{raemaekers.EA:12,  author = {M. Raemaekers and S. {du Plessis} and N.F. Ramsey  and J.M.H. Weusten and M. Vink},  title = {Test–retest variability underlying {fMRI}  measurements},  journal = NI,  year = 2012,  volume = 60,  pages = {717--727},  abstract = {Introduction: A high test–retest reliability is of  pivotal importance for many disciplines in fMRI  research. To assess the current limits of fMRI  reliability, we estimated the variability in true  underlying Blood Oxygen Level Dependent (BOLD)  activation, with which we mean the variability that  would be found in the theoret- ical case when we  could obtain an unlimited number of scans in each  measurement. Methods: In this test–retest study,  subjects were scanned twice with one week apart,  while performing a vi- sual and a motor inhibition  task. We addressed the nature of the variability in  the underlying BOLD signal, by separating for each  brain area and each subject the between-session  differences in the spatial pattern of BOLD  activation, and the global (whole brain) changes in  the amplitude of the spatial pattern of BOLD  activation. Results: We found evidence for changes  in the true underlying spatial pattern of BOLD  activation for both tasks across the two  sessions. The sizes of these changes in pattern  activation were approximately 16\% of the total  activation within the pattern, irrespective of brain  area and task. After spatial smoothing, this vari-  ability was greatly reduced, which suggests it takes  place at a small spatial scale. The mean  between-session differences in the amplitude of  activation across the whole brain were 13.8\% for  the visual task and 23.4\% for the motor inhibition  task. Conclusions: Between-session changes in the  true underlying spatial pattern of BOLD activation  are always present, but occur at a scale that is  consistent with partial voluming effects or spatial  distortions. We found no evidence that the  reliability of the spatial pattern of activation  differs systematically between brain  areas. Consequently, between-session changes in the  amplitude of activation are probably due to global  ef- fects. The observed variability in amplitude  across sessions warrants caution when interpreting  fMRI esti- mates of height of brain activation. A  Matlab implementation of the used algorithm is  available for download at  www.ni-utrecht.nl/downloads/ura.}  }  @Article{raninen.EA:07,  author = {A. Raninen and S. Vanni and L. Hyv{\"a}rinen and  R. N{\"a}s{\"a}nen},  title = {Temporal sensitivity in a hemianopic visual field  can be improved by long-term training using flicker  stimulation},  journal = JNNP,  year = 2007,  volume = 78,  pages = {66--73},  note = {doi:10.1136/jnnp.2006.099366 },  abstract = {Background: Blindness of a visual half-field  (hemianopia) is a common symptom after  postchiasmatic cerebral lesions. Although hemianopia  severely limits activities of daily life, current  clinical practice comprises no training of visual  functions in the blind hemifield. Objective: To  find out whether flicker sensitivity in the blind  hemifield can be improved with intensive training,  and whether training with flicker stimulation can  evoke changes in cortical responsiveness. Methods:  Two men with homonymous hemianopia participated in  the experiments. They trained with flicker stimuli  at 30° or with flickering letters at 10°  eccentricity twice a week for a year, and continued  training with more peripheral stimuli  thereafter. Neuromagnetic responses were registered  at 1–2-month intervals, and the Goldmann perimetry  was recorded before, during and after training.  Results: Flicker sensitivity in the blind hemifield  improved to the level of the intact hemifield within  30° eccentricity in one participant and 20°  eccentricity in the other. Flickering letters were  recognised equally at 10° eccentricity in the blind  and intact hemifields. Improvement spread from the  stimulated horizontal meridian to the whole  hemianopic field within 30°. Before training,  neuromagnetic recordings showed no signal above the  noise level in the hemianopic side. During training,  evoked fields emerged in both participants. No  changes were found in the Goldmann perimetry.  Discussion: Results show that sensitivity to flicker  could be fully restored in the stimulated region,  that improvement in sensitivity spreads to the  surrounding neuronal networks, and that, during  training, accompanying changes occurred in the  neuromagnetic fields. }  }  @Article{raposo.EA:11,  author = {Nicolas Raposo and Alexandra S{\'e}verac Cauquil and  Jean Fran{\,c}ois Albucher and Blandine Acket and  Simona Celebrini and J{\'e}r{\'e}mie Pariente and  Christophe Cognard and Isabelle Loubinoux and  Fran{\,c}ois Chollet},  title = {Poststroke conscious visual deficit: clinical course  and changes in cerebral activations},  journal = {Neurorehabil Neural Repair},  year = 2011,  volume = 25,  number = 8,  pages = {703--710},  annote = {BACKGROUND AND PURPOSE: Little is known about the  outcome and recovery mechanisms of visual perception  after a focal lesion of the occipital lobe in  humans, especially after stroke. In this study, the  authors aimed to describe the clinical course and  the neural substrates of conscious perceptive visual  deficit after posterior cerebral artery infarct.  METHODS: The authors prospectively included 8  patients (7 men and 1 woman; mean age, 64.6 ± 18  years) with visual deficit induced by partial damage  of the striate cortex related to acute posterior  cerebral artery infarct. Conscious perception of  color and motion was assessed from the acute phase  to the third month. Functional magnetic resonance  imaging was performed to investigate neural  substrates of visual recovery. RESULTS: In the  acute phase of stroke, visual deficiency was global  (3/8 patients), selective to color (4/8 patients),  or selective to motion (1/8 patients). During the  follow-up, visual performance increased with respect  to color (from 29\% to 70\%; P < .005) and with  respect to motion (from 47\% to 74\%; P <  .005). Despite a lack of ipsilesional V1 area  activation in the acute phase, activations in this  area and in the contralesional extrastriate cortex  were obtained during follow-up. Both ipsilesional  and contralesional V4 activations were correlated  with better outcome. CONCLUSIONS: Extensive visual  recovery occurs early after partial acute posterior  cerebral artery infarct. Spared islands in  ipsilesional V1 area and transcallosal pathways  might be involved in poststroke visual recovery.}  }  @Article{regan.spekreijse:86,  author = {D. Regan and H. Spekreijse},  title = {Evoked potentials in vision research 1961--86},  journal = VisRes,  year = 1986,  volume = 26,  pages = {1461--1480}  }  @Article{regan:09,  author = {D. Regan},  title = {Some early uses of evoked brain responses in  investigations of human visual function},  journal = VisRes,  year = 2009,  volume = 49,  pages = {882--897},  abstract = {In the context of the technical possibilities of the  time, this paper describes early attempts to employ  visual-evoked potentials (VEPs) as a tool for  investigating human visual function, focussing on  the con- tributions of Henk Spekreijse and his  colleagues. The topics covered are as follows:  attempts to distin- guish between true  pattern-specific VEPs and the effects of responses  to changes in local luminance; retinal rivalry and  interocular sustained suppression; the implications  of VEPs elicited by equiluminant chromatic patterns;  VEPs specific to real and apparent motion; stereo  VEPs; identification of a visual– auditory  convergence area in the human brain.}  }  @Article{reitsma.EA:13,  author = {Danielle C. Reitsma and Jedidiah Mathis and John  L. Ulmer and Wade Mueller and Mary J. Maciejewski  and Edgar A. DeYoe},  title = {Atypical retinotopic organization of visual cortex  in patients with central brain damage: congenital  and adult onset},  journal = JoN,  year = 2013,  volume = 33,  number = 32,  pages = {13010--13024},  abstract = {It remains unclear to what extent retinotopic maps  can undergo large-scale plasticity following damage  to human visual cortex. The literature has  predominately focused on retinotopic changes in  patients with retinal pathologies or congenital  brain malformations. Yet, damage to the adult visual  cortex itself is common in cases such as stroke,  tumor, or trauma. To address this issue, we used a  unique database of fMRI vision maps in patients with  adult-onset (n=25) and congenital (n=2) pathology of  the visual cortex. We identified atypical  retinotopic organization in three patients (two with  adult-onset, and one with congenital pathology)  consisting of an expanded ipsilateral field  representation that was on average 3.2 times greater  than healthy controls. The expanded representations  were located at the vertical meridian borders  between visual areas such as V1/V2. Additionally,  two of the three patients had apparently an ectopic  (topographically inconsistent) representation of the  ipsilateral field within lateral occipital cortex  that is normally associated with visual areas V3/V3A  (and possibly other areas). Both adult-onset cases  had direct damage to early visual cortex itself  (rather than to the afferent drive only), resulting  in a mostly nonfunctional hemisphere. The congenital  case had severe cortical malformation of the visual  cortex and was acallosal. Our results are consistent  with a competitive model in which unilateral damage  to visual cortex or disruption of the transcallosal  connections removes interhemispheric suppression  from retino-geniculate afferents in intact visual  cortex that represent the vertical meridian and  ipsilateral visual field.}  }  @Article{ritchie.EA:12,  author = {Kay L. Ritchie and Amelia R. Hunt and Arash  Sahraie},  title = {Trans-saccadic priming in hemianopia: Sighted-field  sensitivity is boosted by a blind-field prime},  journal = {Neuropsychologia},  year = 2012,  volume = 50,  pages = {997--1005},  abstract = {We experience visual stability despite shifts of the  visual array across the retina produced by eye move-  ments. A process known as remapping is thought to  keep track of the spatial locations of objects as  they move on the retina. We explored remapping in  damaged visual cortex by presenting a stimulus in  the blind field of two patients with  hemianopia. When they executed a saccadic eye  movement that would bring the stimulated location  into the sighted field, reported awareness of the  stimulus increased, even though the stimulus was  removed before the saccade began and so never  actually fell in the sighted field. Moreover, when a  location was primed by a blind-field stimulus and  then brought into the sighted field by a saccade,  detection sensitivity for near-threshold targets  appearing at this location increased  dramatically. The results demonstrate that brain  areas supporting conscious vision are not necessary  for remapping, and suggest visual stability is  maintained for salient objects even when they are  not consciously perceived.}  }  @Article{rossion.EA:12,  author = {Bruno Rossion and Esther Alonso Prieto an Adriano  Boremanse and Dana Kuefner and Goedele {Van Belle}},  title = {A steady-state visual evoked potential approach to  individual face perception: Effect of inversion,  contrast-reversal and temporal dynamics},  journal = NI,  year = 2012,  volume = 63,  pages = {1585--1600},  abstract = {Presentation of a face stimulus for several seconds  at a periodic frequency rate leads to a right  occipito-temporal evoked steady-state visual  potential (SSVEP) confined to the stimulation  frequency band. According to recent evidence  (Rossion and Boremanse, 2011), this face-related  SSVEP is largely reduced in amplitude when the exact  same face is repeated at every stimulation cycle as  compared to the presentation of different individual  faces. Here this SSVEP individual face repetition  effect was tested in 20 participants stimulated with  faces at a 4 Hz rate for 84 s, in 4 conditions:  faces upright or inverted, normal or  contrast-reversed (2 × 2 design). To study the  temporal dynamics of this effect, all stimulation  sequences started with 15 s of identical faces,  after which, in half of the sequences, different  faces were introduced. A larger response to  different than identical faces at the fundamental (4  Hz) and second harmonic (8 Hz) components was  observed for upright faces over the right  occipito-temporal cortex. Weaker effects were found  for inverted and contrast-reversed faces, two  stimulus manipulations that are known to greatly  affect the perception of facial identity. Addition  of the two manipulations further decreased the  effect. The phase of the fundamental frequency SSVEP  response was delayed for inverted and  contrast-reversed faces, to the same extent as the  latency delay observed at the peak of the  face-sensitive N170 component observed at  stimulation sequence onset. Time-course analysis of  the entire sequence of stimulation showed an  immediate increase of 4 Hz amplitude at the onset  (16th second) of different face presentation,  indicating a fast, large and frequency-specific  release to individual face adaptation in the human  brain. Altogether, these observations increase our  understanding of the characteristics of the human  steady-state face potential response and provide  further support for the interest of this approach in  the study of the neurofunctional mechanisms of face  perception.}  }  @Article{sahraie.EA:97,  author = {A. Sahraie and L. Weiskrantz and J. L. Barbur and  A. Simmons and S. C. R. Williams and M. J. Brammer},  title = {Pattern of neuronal activity associated with  conscious and unconscious processing of  visual signals},  journal = PNAS,  year = 1997,  volume = 94,  number = 17,  pages = {9406--9411},  abstract = {Following striate cortex damage in monkeys and  humans there can be residual function mediated by  parallel visual pathways. In humans this can  sometimes be associated with a “feeling” that  something has happened, especially with rapid  movement or abrupt onset. For less transient events,  discriminative performance may still be well above  chance even when the subject reports no conscious  awareness of the stimulus. In a previous study we  examined parameters that yield good residual visual  performance in the “blind” hemifield of a subject  with unilateral damage to the primary visual  cortex. With appropriate parameters we demonstrated  good discriminative performance, both with and  without conscious awareness of a visual event. These  observations raise the possibility of imaging the  brain activity generated in the “aware” and the  “unaware” modes, with matched levels of  discrimination performance, and hence of revealing  patterns of brain activation associated with visual  awareness. The intact hemifield also allows a  comparison with normal vision. Here we report the  results of a functional magnetic resonance imaging  study on the same subject carried out under aware  and unaware stimulus conditions. The results point  to a shift in the pattern of activity from neocortex  in the aware mode, to subcortical structures in the  unaware mode. In the aware mode prestriate and  dorsolateral prefrontal cortices (area 46) are  active. In the unaware mode the superior colliculus  is active, together with medial and orbital  prefrontal cortical sites. }  }  @Article{sammer.EA:05,  author = {Gebhard Sammer and Carlo Blecker and Helge Gebhardt  and Peter Kirsch and Rudolf Stark and Dieter Vaitl},  title = {Acquisition of typical {EEG} waveforms during {fMRI:  SSVEP, LRP}, and frontal theta},  journal = NI,  year = 2005,  volume = 24,  pages = {1012--1024},  abstract = {Recent work has demonstrated the feasibility of  simultaneous electro- encephalography (EEG) and  functional magnetic resonance imaging  (fMRI). Virtually no systematic comparisons between  EEG recorded inside and outside the MR scanner have  been conducted, and it is unknown if different kinds  of frequency mix, topography, and domain- specific  processing are uniformly recordable within the  scanner environment. The aim of the study was to  investigate several typical EEG waveforms in the  same subjects inside the magnet during fMRI and  outside the MR examination room. We examined whether  uniform artifact subtraction allows the extraction  of these different EEG waveforms inside the scanner  during EPI scanning to the same extent as outside  the scanner. Three well-established experiments were  conducted, eliciting steady state visual evoked  potentials (SSVEP), lateralized readiness potentials  (LRP), and frontal theta enhancement induced by  mental addition. All waveforms could be extracted  from the EEG recorded during fMRI. Substantially no  differences in these waveforms of interest were  found between gradient-switching and intermediate  epochs during fMRI (only the SSVEP-experiment was  designed for a comparison of gradient—with  intermediate epochs), or between waveforms recorded  inside the scanner during EPI scanning and outside  the MR examination room (all experiments). However,  non-specific amplitude differences were found  between inside and outside recorded EEG at lateral  electrodes, which were not in any interaction with  the effects of interest. The source of these  differences requires further exploration. The high  concordance of activation patterns with published  results demonstrates that EPI-images could be  acquired during EEG recording without significant  distortion.}  }  @Article{sanchezpanchuelo.EA:12,  author = {Rosa M. S{\'a}nchez-Panchuelo and Susan T. Francis  and Denis Schluppeck and Richard W. Bowtell},  title = {Correspondence of Human Visual Areas Identified  Using Functional and Anatomical {MRI} In Vivo at {7  T}},  journal = JMRI,  year = 2012,  volume = 35,  pages = {287--299},  abstract = {Purpose: To study the correspondence of anatomically  and functionally defined visual areas (primary  visual cor- tex, V1, and motion selective area  V5/human MTþ) by using structural magnetic resonance  imaging (MRI) and functional MRI (fMRI) in vivo at 7  T. Materials and Methods: Four subjects  participated in this study. High-resolution (\%0.4 mm  isotropic) anatomi- cal MRI was used to identify  cortical regions based on their distinct cortical  lamination. The optimal contrast for identifying  heavily myelinated layers within gray matter was  quantitatively assessed by comparing T1-weighted  magnetization-prepared rapid gradient echo (MPRAGE)  and T2*-weighted, 3D fast-low angle shot (FLASH)  imaging. Retinotopic mapping was performed using  GE-based fMRI at 1.5 mm isotropic resolution to  identify functional areas. Results: T2*-weighted  FLASH imaging was found to pro- vide a significantly  higher contrast-to-noise ratio, allowing  visualization of the stria of Gennari in every slice  of a vol- ume covering the occipital cortex in each  of the four sub- jects in this study. The  independently derived boundary of V1, identified in  the same subjects using retinotopic mapping by fMRI,  closely matched the border of anatomi- cally defined  striate cortex in the human brain. Evidence of  banding was also found within the functionally  defined V5 area; however, we did not find a good  correlation of this area, or the functionally  identified subregion (MT), with the banded area.  Conclusion: High-resolution T2*-weighted images  acquired at 7 T can be used to identify myelinated  bands within cortical gray matter in reasonable  measurement times. Regions where a myelinated band  was identified show a high degree of overlap with  the functionally defined V1 area.}  }  @Article{savazzi.marzi:02,  author = {Silvia Savazzi and Carlo A. Marzi},  title = {Speeding Up Reaction Time with Invisible Stimuli},  journal = CurrBio,  year = 2002,  volume = 12,  pages = {403--407},  abstract = {Normal subjects react more quickly to a pair of  visual stimuli than to a stimulus alone. This  phenomenon is known as the redundant signal effect  (RSE) [1] and represents an example of divided  visual attention in which signal processing is  carried out in parallel to the advantage of response  speed. A most interesting aspect of this phenomenon  is that it can occur when one stimulus in a pair  cannot be consciously detected because of hemianopia  [2, 3] or unilateral extinction [4] resulting from  brain damage. Here, we report that a similar  dissociation between visual awareness and visually  guided behavior is present in normal subjects who  show an RSE even when the luminance of one of a pair  of stimuli is below detection threshold. The  observed RSE cannot be attributed to probability  sum- mation because it violates Miller’s race  inequality [5] and is likely to be related to neural  summation between supra- and subthreshold  stimuli. Given that a similar implicit RSE is  present in hemispherectomy patients [3], we  hypothesize that the site of this summation might be  the superior colliculus (SC).}  }  @Article{savazzi.EA:07,  author = {Silvia Savazzi and Mara Fabri and Guido Rubboli and  Aldo Paggiand Carlo Alberto Tassinari and Carlo  Alberto Marzi},  title = {Interhemispheric transfer following callosotomy in  humans: Role of the superior colliculus},  journal = {Neuropsychologia},  year = 2007,  volume = 45,  pages = {2417--2427},  abstract = {It is now common knowledge that the total surgical  section of the corpus callosum (CC) and of the other  forebrain commissures prevents interhemispheric  transfer (IT) of a host of mental functions. By  contrast, IT of simple sensorimotor functions,  although severely delayed, is not abolished, and an  important question concerns the pathways subserving  this residual IT. To answer this question we  assessed visuomotor IT in split-brain patients using  the Poffenberger paradigm (PP), that is, a  behavioral paradigm in which simple reaction time  (RT) to visual stimuli presented to the hemifield  ipsilateral to the responding hand is compared to  stimuli presented to the contralateral hemifield, a  condition requiring an IT. We tested the possibility  that the residual IT is mediated by the collicular  commissure interconnecting the two sides of the  superior colliculus (SC). To this purpose, we used  short-wavelength visual stimuli, which in  neurophysiological studies in non-human primates  have been shown to be undetectable by collicular  neurons. We found that, in both totally and  partially callosotomised patients, IT was  considerably longer with S-cone input than with  L-cone input or with achromatic stimuli. This was  not the case in healthy participants in whom IT was  not affected by color. These data clearly show that  the SC plays an important role in IT of sensorimotor  information in the absence of the corpus callosum.}  }  @Article{savoy:05,  author = {Robert L. Savoy},  title = {Experimental design in brain activation {MRI}:  Cautionary tales},  journal = BRB,  year = 2005,  volume = 67,  pages = {361--367},  abstract = {The use of functional magnetic resonance imaging  (fMRI) in cognitive neuroscience has expanded at an  amazing rate in the past 10 years. Current research  includes increasingly subtle and specific attempts  to dissect the cognitive and emotional mechanisms  called into play when humans make decisions. The  present essay will briefly review some of the  general considerations and domains of information  needed when one designs fMRI-based  experiments. However, the main theme will be the  difficulties associated with designing, conducting,  analyzing and interpreting such research. Functional  MRI is an unusually complicated technique, and there  are numerous ways for experiments to go wrong. As  well as demanding exceptional care in maintaining  the quality of one’s own research, this makes the  universal problem of evaluating other peoples’  research particularly challenging.}  }  @Article{savoy:06,  author = {Robert L. Savoy},  title = {Using Small Numbers of Subjects in {fMRI}-Based  Research},  journal = IEEEmb,  year = 2006,  volume = 25,  pages = {52--59},  abstract = {Most discussions of brain activation functional mag-  netic resonance imaging (fMRI) data start with the  analysis of images from a single subject but move  rather quickly to the issues associated with group  comparisons and drawing population  conclusions. There are many technical issues  associated with those group compar- isons that do  not arise, or are very different, for data from a  single subject. In addition, there are contexts  (most notably clinical evaluations) where studying  an individual subject is essential. This article  reviews a number of issues related to the special  features and potentials associated with doing  function- al brain imaging experiments with one or a  small number of subjects. It concludes with the  speculation that, ironically, it may someday turn  out that the information from a few brains,  thoroughly studied, will reveal more about the  universal aspects of human brain function and  organization than the cur- rent torrent of studies  from large collections of brains.}  }  @Article{scarpazza.EA:13,  author = {C. Scarpazza and G. Sartori and M.S. {De Simone} and  A. Mechelli},  title = {When the single matters more than the group: Very  high false positive rates in single case Voxel Based  Morphometry},  journal = NI,  year = 2013,  volume = 70,  pages = {175--188},  abstract = {Voxel Based Morphometry (VBM) studies typically  involve a comparison between groups of individuals;  this ap- proach however does not allow inferences to  be made at the level of the individual. In recent  years, an increasing number of research groups have  attempted to overcome this issue by performing  single case studies, which in- volve the comparison  between a single subject and a control  group. However, the interpretation of the results is  problematic; for instance, any significant  difference might be driven by individual variability  in neuroanatomy rather than the neuropathology of  the disease under investigation, or might represent  a false positive due to the data being sampled from  non-normally distributed populations. The aim of the  present investigation was to empirically estimate  the likelihood of detecting significant differences  in gray matter volume in individuals free from  neurological or psychiatric diagnosis. We compared a  total of 200 single subjects against a group of 16  controls matched for age and gender, using two  independent datasets from the Neuroimaging  Informatics Tools and Resources Clearinghouse. We  report that the chance of detecting a significant  difference in a disease-free individual is much  higher than previously expected; for instance, using  a standard voxel-wise threshold of p b 0.05  (corrected) and an extent threshold of 10 voxels,  the likelihood of a single subject showing at least  one significant difference is as high as 93.5\% for  increases and 71\% for decreases. We also report  that the chance of detecting significant differences  was greatest in frontal and temporal cortices and  lowest in subcortical regions. The chance of  detecting significant differences was inversely  related to the degree of smoothing applied to the  data, and was higher for unmodulated than modulated  data. These results were replicated in the two in-  dependent datasets. By providing an empirical  estimation of the number of significant increases  and decreases to be expected in each cortical and  subcortical region in disease-free individuals, the  present investigation could in- form the  interpretation of future single case VBM studies.}  }  @Article{schmid.EA:13,  author = {Michael C. Schmid and Joscha T. Schmiedt and Andrew  J. Peters2and Richard C. Saunders and Alexander  Maier and David A. Leopold},  title = {Motion-Sensitive Responses in Visual Area {V4} in  the Absence of Primary Visual Cortex},  journal = JoN,  year = 2013,  volume = 33,  number = 48,  pages = {18740--18745},  abstract = {Neurons in cortical ventral-stream area V4 are  thought to contribute to important aspects of visual  processing by integrating information from primary  visual cortex (V1). However, how V4 neurons respond  to visual stimulation after V1 injury remains  unclear: While electrophysiological investigation of  V4 neurons during reversible V1 inactivation  suggests that virtually all responses are eliminated  (Girard et al., 1991), fMRI in humans and monkeys  with permanent lesions shows reliable V1-independent  activity (Baseler et al., 1999; Goebel et al., 2001;  Schmid et al., 2010). To resolve this apparent  discrepancy, we longitudinally assessed neuronal  functions of macaque area V4 using chronically  implanted electrode arrays before and after creating  a permanent aspiration lesion in V1. During the  month after lesioning, we observed weak yet  significant spiking activity in response to stimuli  presented to the lesion-affected part of the visual  field. These V1-independent responses showed  sensitivity for motion and likely reflect the effect  of V1-bypassing geniculate input into extrastriate  areas. }  }  @Article{schmidt.EA:02,  author = {Brandon K. Schmidt and Edward K. Vogel and Geoffrey  F. Woodman and Steven J. Luck},  title = {Voluntary and automatic attentional control of  visual working memory},  journal = PP,  year = 2002,  volume = 64,  number = 5,  pages = {754--763},  abstract = {Previous studies of attention-directing cues have  focused largely on the effects of cuing on  perceptual processes, but cuing may also influence  the transfer of perceptual representations into  visual working memory. In the present study, we  examined this potential role of cues, using both  predictive and nonpredictive cues in the context of  a visual working memory task. Each trial began with  a cue, followed by an array of six colored squares,  a delay interval, and then a probe square presented  at the location of one of the squares in the  previous array. The subjects were required to  indicate whether the color of the probe square was  the same as the color of the square that had  previously been presented at the same  location. Performance on this working memory task  was more accurate when the cued location was probed  than when an uncued location was probed, even when  the cued location was no more likely to be probed  than any of the uncued locations. An additional  experiment using the abrupt-onset paradigm of Yantis  and Jonides (1984) yielded similar results. Thus,  visual transients may automatically influence the  transfer of perceptual representations into visual  working memory.}  }  @Article{schoenfeld.EA:02,  author = {Mircea Ariel Schoenfeld and Toemme Noesselt and  Dorothe Poggel and Claus Tempelmann and Jens-Max  Hopf and Martin G. Woldorff and Hans-Jochen Heinze  and Steven A. Hillyard},  title = {Analysis of pathways mediating preserved vision  after striate cortex lesions},  journal = AN,  year = 2002,  volume = 52,  number = 6,  pages = {814--824},  abstract = {This study investigated the neural substrates of  preserved visual functioning in a patient with  homonymous hemianopsia and Riddoch syndrome after a  posterior cerebral artery stroke affecting the  primary visual cortex (area V1). The limited visual  abilities of this patient included above-chance  verbal reports of movement and color change as well  as discrimination of movement direction in the  hemianopic field. Functional magnetic resonance  imaging showed that motion and color-change stimuli  presented to the hemianopic field produced  activation in several extrastriate areas of the  lesioned hemisphere that were defined using  retinotopic mapping. Magnetoencephalographic  recordings indicated that evoked activity occurred  earlier in the higher-tier visual areas V4/V8 and V5  than in the lower-tier areas V2/V3 adjacent to the  lesion. In addition, the functional magnetic  resonance imaging analysis showed an increased  functional connectivity between areas V4/V8 and V5  of the lesioned hemisphere in comparison with the  same areas in the intact hemisphere during the  presentation of color changes. These results suggest  that visual perception after the V1 lesion in  Riddoch syndrome is mediated by subcortical pathways  that bypass V1 and project first to higher-tier  visual areas V5 and V4/V8 and subsequently to  lower-tier areas V2/V3.}  }  @Article{silvanto:15,  author = {Juha Silvanto},  title = {Why is ``blindsight'' blind? A new perspective on  primary visual cortex, recurrent activity and visual  awareness},  journal = ConscCog,  year = 2015,  volume = 32,  pages = {15--32},  abstract = {The neuropsychological phenomenon of blindsight has  been taken to suggest that the pri- mary visual  cortex (V1) plays a unique role in visual awareness,  and that extrastriate acti- vation needs to be fed  back to V1 in order for the content of that  activation to be consciously perceived. The aim of  this review is to evaluate this theoretical  framework and to revisit its key tenets. Firstly, is  blindsight truly a dissociation of awareness and  visual detection? Secondly, is there sufficient  evidence to rule out the possibility that the loss  of awareness resulting from a V1 lesion simply  reflects reduced extrastriate respon- siveness,  rather than a unique role of V1 in conscious  experience? Evaluation of these argu- ments and the  empirical evidence leads to the conclusion that the  loss of phenomenal awareness in blindsight may not  be due to feedback activity in V1 being the hallmark  awareness. On the basis of existing literature, an  alternative explanation of blindsight is  proposed. In this view, visual awareness is a  ‘‘global’’ cognitive function as its hallmark is the  availability of information to a large number of  perceptual and cognitive systems; this requires  inter-areal long-range synchronous oscillatory  activity. For these oscillations to arise, a  specific temporal profile of neuronal activity is  required, which is established through recurrent  feedback activity involving V1 and the extrastriate  cortex. When V1 is lesioned, the loss of recurrent  activity prevents inter-areal networks on the basis  of oscil- latory activity. However, as limited  amount of input can reach extrastriate cortex and  some extrastriate neuronal selectivity is preserved,  computations involving comparison of neural firing  rates within a cortical area remain possible. This  enables ‘‘local’’ read-out from spe- cific brain  regions, allowing for the detection and  discrimination of basic visual attributes. Thus  blindsight is blind due to lack of ‘‘global’’  long-range synchrony, and it functions via ‘‘local’’  neural readout from extrastriate areas.}  }  @Article{silvanto.EA:09,  author = {Juha Silvanto and Vincent Walsh and Alan Cowey},  title = {Abnormal functional connectivity between  ipsilesional {V5/MT+} and contralesional striate  cortex {(V1)} in blindsight},  journal = EBR,  year = 2009,  volume = 193,  number = 4,  pages = {645--650},  abstract = {Damage to the visual cortex can lead to changes in  anatomical connectivity between the remaining  areas. For example, after a unilateral lesion to  striate cortex (V1), an abnormal anatomical pathway  can develop between the lateral geniculate nucleus  of the undamaged hemisphere and the motion area  V5/MT+ in the damaged hemisphere, accompanied by a  hypernormal callosal connection between the area  V5/MT+ of the two hemispheres. Here we investigated,  using transcranial magnetic stimulation (TMS), the  functional significance of these pathways in the  blindsight subject GY, in whom they were first  demonstrated. We show that TMS applied over the  extrastriate area V5/MT+ in GY’s damaged hemisphere  modulates the appearance of phosphenes induced from  V1 in the normal hemisphere. In contrast, in  neurologically normal control subjects, TMS applied  over V5/MT+ never influenced the phosphenes induced  from V1 in the other hemisphere. The findings  indicate an abnormal functional connectivity between  V5/MT in the damaged hemisphere and the early visual  cortex in the normal hemisphere, consistent with  GY’s abnormal anatomical connectivity.}  }  @Article{singh.EA:03,  author = {Manbir Singh and Sungheon Kim and Tae-Seong Kim},  title = {Correlation between {BOLD-fMRI and EEG} signal  changes in response to visual stimulus frequency in  humans},  journal = MRM,  year = 2003,  volume = 49,  pages = {108--114},  abstract = {The correlation between signals acquired using  electroencephalography (EEG) and fMRI was  investigated in humans during visual  stimulation. Evoked potential EEG and BOLD fMRI data  were acquired independently under similar conditions  from eight subjects during stimulation by a  checkerboard flashed at frequencies ranging from  2–12 Hz. The results indicate highly correlated  changes in the strength of the EEG signal averaged  over two occipital electrodes and the BOLD signal  within the occipital lobe as a function of flash  frequency for 7/8 subjects (average linear  correlation coefficient of 0.76). Both signals  peaked at approximately 8 Hz. For one subject the  correlation coefficient was 0.20; the EEG signal  peaked at 6 Hz and the BOLD signal peaked at 10  Hz. Overall, the EEG and BOLD signals, each averaged  over 40-sec stimulation periods, appear to be  coupled linearly during visual stimulation by a  flashing checkerboard.}  }  @Article{skudlarski.EA:99,  author = {Pawel Skudlarski and R. Todd Constable and John  C. Gore},  title = {{ROC} Analysis of Statistical Methods Used in  Functional {MRI}: Individual Subjects},  journal = NI,  year = 1999,  volume = 9,  pages = {311--329},  abstract = {The complicated structure of fMRI signals and asso-  ciated noise sources make it difficult to assess the  validity of various steps involved in the  statistical analysis of brain activation. Most  methods used for fMRI analysis assume that  observations are indepen- dent and that the noise  can be treated as white gauss- ian noise. These  assumptions are usually not true but it is difficult  to assess how severely these assumptions are  violated and what are their practical conse-  quences. In this study a direct comparison is made  between the power of various analytical methods used  to detect activations, without reference to  estimates of statistical significance. The  statistics used in fMRI are treated as metrics  designed to detect activations and are not  interpreted probabilistically. The receiver op-  erator characteristic (ROC) method is used to  compare the efficacy of various steps in calculating  an activa- tion map in the study of a single subject  based on optimizing the ratio of the number of  detected activa- tions to the number of  false-positive findings. The main findings are as  follows: Preprocessing. The removal of intensity  drifts and high-pass filtering applied on the voxel  time-course level is beneficial to the efficacy of  analysis. Temporal normalization of the global image  intensity, smoothing in the temporal domain, and  low- pass filtering do not improve power of  analysis. Choices of statistics. the  cross-correlation coefficient and t -statistic, as  well as nonparametric Mann–Whitney statistics, prove  to be the most effective and are similar in  performance, by our criterion. Task design. the  proper design of task protocols is shown to be  crucial. In an alternating block design the optimal  block length is be approximately 18 s. Spatial  clustering. an initial spatial smoothing of images  is more efficient than cluster filtering of the  statistical parametric activation maps.}  }  @Article{smith.EA:09,  author = {A. T. Smith and P. L. Cotton and A. Bruno and  C. Moutsiana},  title = {Dissociating Vision and Visual Attention in the  Human Pulvinar},  journal = JoNeurophys,  year = 2009,  key = 101,  volume = 2,  pages = {917--925},  abstract = {The pulvinar region of the thalamus has repeatedly  been linked with the control of attention. However,  the functions of the pulvinar remain poorly  characterized, both in human and in nonhuman  primates. In a functional MRI study, we examined the  relative contributions to activity in the human  posterior pulvinar made by visual drive (the  presence of an unattended visual stimulus) and  attention (covert spatial attention to the  stimulus). In an event-related design, large optic  flow stimuli were presented to the left and/or right  of a central fixation point. When unattended, the  stimuli robustly activated two regions of the  pulvinar, one medial and one dorsal with respect to  the lateral geniculate. The activity in both regions  shows a strong contralateral bias, suggesting  retinotopic organization. Primate physiology  suggests that the two regions could be two portions  of the same double map of the visual field. In our  paradigm, attending to the stimulus enhanced the  response by about 20\%. Thus attention is not  necessary to activate the human pulvinar and the  degree of attentional enhancement matches, but does  not exceed, that seen in the cortical regions with  which the posterior pulvinar connects.}  }  @Article{srinivasan.EA:07,  author = {Ramesh Srinivasan and Eleonora Fornari and Maria  G. Knyazeva and Reto Meuli and Philippe Maeder},  title = {{fMRI} responses in medial frontal cortex that  depend on the temporal frequency of visual input},  journal = ExpBrRes,  year = 2007,  volume = 180,  number = 4,  pages = {677--691},  abstract = {Functional networks in the human brain have been  investigated using electrophysiological methods  (EEG/MEG, LFP, and MUA) and steady-state paradigms  that apply periodic luminance or contrast modulation  to drive cortical networks. We have used this  approach with fMRI to characterize a cortical  network driven by a checkerboard reversing at a  fixed frequency. We found that the fMRI signals in  voxels located in occipital cortex were increased by  checkerboard reversal at frequencies ranging from 3  to 14 Hz. In contrast, the response of a cluster of  voxels centered on basal medial frontal cortex  depended strongly on the reversal frequency,  consistently exhibiting a peak in the response for  specific reversal frequencies between 3 and 5 Hz in  each subject. The fMRI signals at the frontal voxels  were positively correlated indicating a homogeneous  cluster. Some of the occipital voxels were  positively correlated to the frontal voxels  apparently forming a large-scale functional  network. Other occipital voxels were negatively  correlated to the frontal voxels, suggesting a  functionally distinct network. The results provide  preliminary fMRI evidence that during visual  stimulation, input frequency can be varied to engage  different functional networks.}  }  @Article{sterzer.kleinschmidt:10,  author = {Philipp Sterzer and Andreas Kleinschmidt},  title = {Anterior insula activations in perceptual paradigms:  often observed but barely understood},  journal = BSF,  year = 2010,  volume = 214,  pages = {611--622},  abstract = {Anterior insular cortex is among the non-sen- sory  brain regions most commonly found activated in  functional brain imaging studies on visual and  auditory perception. However, most of these studies  do not explic- itly address the functional role of  this specific brain region in perception, but rather  report its activation as a by- product. Here, we  attempt to characterize the involvement of anterior  insular cortex in various perceptual paradigms,  including studies of visual awareness, perceptual  decision making, cross-modal sensory processes and  the role of spontaneous neural activity fluctuations  in perception. We conclude that anterior insular  cortex may be associated with perception in that it  underpins heightened alertness of either stimulus-  or task-driven origin, or both. Such a mechanism  could integrate endogenous and exogenous functional  demands under the joint criterion of whether they  challenge an individual’s homeostasis.}  }  @Article{stevens.EA:13,  author = {M. Tynan R. Stevens and Ryan C.N. D'Arcye and  Gerhard Stroink and David B. Clarke and Steven  D. Beyea},  title = {Thresholds in {fMRI} studies: Reliable for single  subjects?},  journal = JNMeth,  year = 2013,  volume = 219,  pages = {312--323},  abstract = {Many studies have investigated test–retest  reliability of active voxel classification for fMRI,  which is increasingly important for emerging  clinical applications. The implicit impact of  voxel-wise thresholding on this type of reliability  has previously been under-appreciated. This has had  two detrimental effects: (1) reliability studies use  different fixed thresholds, making comparison of  results is challenging; (2) typical studies do not  assess reliability at the individual level, which  could provide information for selecting acti- vation  thresholds. To show the limitations of traditional  fixed-threshold approaches, we investigated the  threshold dependence of fMRI reliability measures,  with the goal of developing an automated threshold  selection routine. For this purpose, we demonstrated  threshold dependence of both novel (ROC-reliability  or ROC-r) and established (Rombouts overlap or RR)  reliability measures. Both methods rely minimally on  statistical assumptions, and provide a data-driven  summary of the threshold–reliability  relationship. We applied these methods to data from  eight subjects performing a simple finger tapping  task across repeated fMRI sessions. We showed that  the reliability measures varied dramatically with  threshold. This variation depended strongly on the  individual tested. Finally, we demonstrated novel  procedures using ROC-r and overlap analysis to  optimize thresholds on a case-by-case  basis. Ultimately, a method to determine robust  individual-level activation maps represents a  critical advance for fMRI as a diagnostic tool.}  }  @Article{stoerig.cowey:97,  author = {Petra Stoerig and Alan Cowey},  title = {Blindsight in man and monkey},  journal = {Brain},  year = 1997,  volume = 120,  pages = {535--559},  abstract = {In man and monkey, absolute cortical blindness is  caused by degeneration. While extrastriate cortical  areas participate in the mediation of the  forced-choice responses, a concomitant destruction  of the optic radiations and/or the primary visual  cortex. It is characterized by an absence of any  conscious striate cortical activation does not seem  to be necessary for blindsight. Whether the loss of  phenomenal vision is a vision, but stimuli presented  inside its borders may nevertheless be  processed. This unconscious vision includes  necessary consequnce of striate cortical destruction  and whether this structure is indispensable for  conscious sight neuroendocrine, reflexive, indirect  and forced-choice responses which are mediated by  the visual subsystems are much debated questions  which need to be tackled experimentally.}  }  @Article{stoerig.EA:98,  author = {Petra Stoerig and Andreas Kleinschmidt and Jens  Frahm},  title = {No visual responses in denervated V1:  high-resolution functional magnetic resonance  imaging of a blindsight patient},  journal = {Neuroreport},  year = 1998,  volume = 9,  number = 1,  pages = {21--25},  abstract = {Following severe cranio-cerebral trauma that  affected the optic radiation, patient FS suffers  from an incomplete macula-splitting  hemianopia. Within the hemianopic field, FS exhibits  blindsight, i.e. he detects and discriminates visual  stimuli he cannot (consciously) see. We performed  functional magnetic resonance imaging (fMRI) at high  spatial resolution using a large flickering stimulus  field to assess visual responsiveness of  deafferented V1. Contrasting strong activation of  the normal contralesional visual cortex,  ipsilesional V1 displayed no stimulus-related MRI  signal changes. However, activation was observed in  ipsilesional extrastriate cortex. We conclude that  blindsight does not depend on functional islands of  tissue preserved within the deafferented striate  cortex.}  }  @Article{tinelli.EA:13,  author = {Francesca Tinelli and Guido Marco Cicchini and  Roberto Arrighi and Michela Tosetti and Giovanni  Cioni and Maria Concetta Morrone},  title = {Blindsight in children with congenital and acquired  cerebral lesions},  journal = {Cortex},  year = 2013,  volume = 49,  pages = {1636--1647},  abstract = {It has been shown that unconscious visual function  can survive lesions to optical radia- tions and/or  primary visual cortex (V1), a phenomenon termed  “blindsight”. Studies on animal models (cat and  monkey) show that the age when the lesion occurs  determines the extent of residual visual  capacities. Much less is known about the functional  and under- lying neuronal repercussions of early  cortical damage in humans. We measured sensitivity  to several visual tasks in four children with  congenital unilateral brain lesions that severely  affected optic radiations, and in another group of  three children with similar lesions, acquired in  childhood. In two of the congenital patients, we  measured blood oxygenation level dependent (BOLD)  activity in response to stimulation of each visual  field quadrants. Results show clear evidence of  residual unconscious processing of position,  orientation and motion of visual stimuli displayed  in the scotoma of congenitally lesioned children,  but not in the children with acquired lesions. The  calcarine cortical BOLD responses were abnormally  elicited by stimulation of the ipsilateral visual  field and in the scotoma region, demonstrating a  profound neuronal reorganization. In conclusion, our  data suggest that congenital lesions can trigger  massive reorganization of the visual system to  alleviate functional effects of early brain  insults.}  }  @Article{toffanin.EA:09,  author = {Paolo Toffanin and Ritske {de Jong} and Addie  Johnson and Sander Martens},  title = {Using frequency tagging to quantify attentional  deployment in a visual divided attention task},  journal = IJP,  year = 2009,  volume = 72,  pages = {289--298},  abstract = {Frequency tagging is an EEG method based on the  quantification of the steady state visual evoked  potential (SSVEP) elicited from stimuli which  flicker with a distinctive frequency. Because the  amplitude of the SSVEP is modulated by attention  such that attended stimuli elicit higher SSVEP  amplitudes than do ignored stimuli, the method has  been used to investigate the neural mechanisms of  spatial attention. However, up to now it has not  been shown whether the amplitude of the SSVEP is  sensitive to gradations of attention and there has  been debate about whether attention effects on the  SSVEP are dependent on the tagging frequency  used. We thus compared attention effects on SSVEP  across three attention conditions—focused, divided,  and ignored—with six different tagging  frequencies. Participants performed a visual  detection task (respond to the digit 5 embedded in a  stream of characters). Two stimulus streams, one to  the left and one to the right of fixation, were  displayed simultaneously, each with a background  grey square whose hue was sine-modulated with one of  the six tagging frequencies. At the beginning of  each trial a cue indicated whether targets on the  left, right, or both sides should be responded  to. Accuracy was higher in the focused- than in the  divided- attention condition. SSVEP amplitudes were  greatest in the focused-attention condition,  intermediate in the divided-attention condition, and  smallest in the ignored-attention condition. The  effect of attention on SSVEP amplitude did not  depend on the tagging frequency used. Frequency  tagging appears to be a flexible technique for  studying attention.}  }  @Article{urbanski.EA:14,  author = {Marika Urbanski and Olivier A. Coubard and Cl{\'e}mence  Bourlon},  title = {Visualizing the blind brain: brain imaging of visual  field defects from early recovery to rehabilitation  techniques},  journal = FIN,  year = 2014,  volume = 8,  number = 74,  note = {doi:10.3389/fnint.2014.00074},  abstract = {Visual field defects (VFDs) are one of the most  common consequences observed after brain injury,  especially after a stroke in the posterior cerebral  artery territory. Less frequently, tumors, traumatic  brain injury, brain surgery or demyelination can  also determine various visual disabilities, from a  decrease in visual acuity to cerebral  blindness. Visual field defects is a factor of bad  functional prognosis as it compromises many daily  life activities (e.g., obstacle avoidance, driving,  and reading) and therefore the patient’s quality of  life. Spontaneous recovery seems to be limited and  restricted to the first 6 months, with the best  chance of improvement at 1 month. The possible  mechanisms at work could be partly due to cortical  reorganization in the visual areas (plasticity)  and/or partly to the use of intact alternative  visual routes, first identified in animal studies  and possibly underlying the phenomenon of  blindsight. Despite processes of early recovery,  which is rarely complete, and learning of  compensatory strategies, the patient’s autonomy may  still be compromised at more chronic  stages. Therefore, various rehabilitation therapies  based on neuroanatomical knowledge have been  developed to improve VFDs. These use eye-movement  training techniques (e.g., visual search, saccadic  eye movements), reading training, visual field  restitution (the Vision Restoration Therapy, VRT),  or perceptual learning. In this review, we will  focus on studies of human adults with acquired VFDs,  which have used different imaging techniques  (Positron Emission Tomography, PET; Diffusion Tensor  Imaging, DTI; functional Magnetic Resonance Imaging,  fMRI; Magneto Encephalography, MEG) or  neurostimulation techniques (Transcranial Magnetic  Stimulation, TMS; transcranial Direct Current  Stimulation, tDCS) to show brain activations in the  course of spontaneous recovery or after specific  rehabilitation techniques.}  }  @InCollection{vallar:00,  author = {G. Vallar},  title = {The methodological foundations of human  neuropsychology: studies in brain-damaged patient},  booktitle = {Handbook of neuropsychology},  publisher = {Elsevier},  year = 2000,  editor = {F. Boller and J. Grafman},  volume = 1,  pages = {53--76},  edition = {2nd},  address = {Amsterdam}  }  @Article{vialatte.EA:10,  author = {Fran{\,c}ois-Beno{\^i}t Vialatte and Monique Maurice  and Justin Dauwels and Andrzej Cichocki},  title = {Steady-state visually evoked potentials: Focus on  essential paradigms and future perspectives},  journal = PN,  year = 2010,  volume = 90,  pages = {418--438},  abstract = {After 40 years of investigation, steady-state  visually evoked potentials (SSVEPs) have been shown  to be useful for many paradigms in cognitive (visual  attention, binocular rivalry, working memory, and  brain rhythms) and clinical neuroscience (aging,  neurodegenerative disorders, schizophrenia,  ophthalmic pathologies, migraine, autism,  depression, anxiety, stress, and  epilepsy). Recently, in engineering, SSVEPs found a  novel application for SSVEP-driven brain–computer  interface (BCI) systems. Although some SSVEP  properties are well documented, many questions are  still hotly debated. We provide an overview of  recent SSVEP studies in neuroscience (using  implanted and scalp EEG, fMRI, or PET), with the  perspective of modern theories about the visual  pathway. We investigate the steady-state evoked  activity, its properties, and the mechanisms behind  SSVEP generation. Next, we describe the SSVEP-BCI  paradigm and review recently developed SSVEP-based  BCI systems. Lastly, we outline future research  directions related to basic and applied aspects of  SSVEPs.}  }  @Article{walters.EA:03,  author = {Nathan B. Walters and Gary F. Egan and Jillian  J. Kril and Michael Kean and Patricia Waley and Mark  Jenkinson and John D. G. Watson},  title = {In vivo identification of human cortical areas using  high-resolution {MRI}: An approach to cerebral  structure–function correlation},  journal = PNAS,  year = 2003,  volume = 100,  number = 5,  pages = {2981--2986},  abstract = {Understanding the relationship between the  structural and func- tional organization of the  human brain is one of the most impor- tant goals of  neuroscience. Individual variability in brain  structure means that it is essential to obtain this  information from the same subject. To date, this has  been almost impossible. Even though noninvasive  functional imaging techniques such as functional MRI  (fMRI) are now commonplace, there is no  complementary nonin- vasive structural technique. We  present an in vivo method of examining the detailed  neuroanatomy of any individual, which can then be  correlated with that individual’s own functional  results. This method utilizes high-resolution  structural MRI to identify distinct cortical regions  based on cortical lamination structure. We  demonstrate that the observed MR lamination patterns  relate to myeloarchitecture through a correlation of  histology with MRI. In vivo high-resolution MRI  studies identify striate cortex, as well as visual  area V5, in four individuals, as defined by using  fMRI. The anatomical identification of a cortical  area (V5 MT) outside of striate cortex is a  significant advance, proving it possible to identify  extra-striate cortical areas and demonstrating that  in vivo struc- tural mapping of the human cerebral  cortex is possible.}  }  @Article{wan.EA:06,  author = {Xiaohong Wan and Jorge Riera and Kazuki Iwata and  Makoto Takahashi and Toshio Wakabayashi and Ryuta  Kawashima},  title = {The neural basis of the hemodynamic response  nonlinearity in human primary visual cortex:  Implications for neurovascular coupling mechanism},  journal = NI,  year = 2006,  volume = 32,  number = 2,  pages = {616--625},  abstract = {It has been well recognized that the nonlinear  hemodynamic responses of the blood oxygenation  level-dependent (BOLD) functional MRI (fMRI) are  important and ubiquitous in a series of experimental  paradigms, especially for the event-related  fMRI. Although this phenomenon has been intensively  studied and it has been found that the  post-capillary venous expansion is an intrinsically  nonlinear mechanical process, the existence of an  additional neural basis for the nonlinearity has not  been clearly shown. In this paper, we assessed the  correlation between the electric and vascular  indices by performing simultaneous  electroencephalography (EEG) and fMRI recordings in  humans during a series of visual stimulation (i.e.,  radial checkerboard). With changes of the visual  stimulation frequencies (from 0.5 to 16 Hz) and  contrasts (from 1% to 100%), both the event related  potentials (ERPs) and hemodynamic responses show  nonlinear behaviors. In particular, the mean power  of the brain electric sources and the neuronal  efficacies (as originally defined in the  hemodynamics model [Friston et al. Neuroimage, 12,  466–477, 2000], here represent the vascular inputs)  in primary visual cortex consistently show a linear  correlation for all subjects. This indicates that  the hemodynamic response nonlinearity found in this  paper primarily reflects the nonlinearity of  underlying neural activity. Most importantly, this  finding underpins a nonlinear neurovascular  coupling. Specifically, it is shown that the  transferring function of the neurovascular coupling  is likely a power transducer, which integrates the  fast dynamics of neural activity into the vascular  input of slow hemodynamics.}  }  @Article{weil.EA:09,  author = {R.S. Weil and G.T. Plant and M. James-Galton and  G. Rees},  title = {Neural correlates of hemianopic completion across  the vertical meridian},  journal = {Neuropsychologia},  year = 2009,  volume = 47,  pages = {457--464},  abstract = {Hemianopic completion refers to the perceptual  completion of figures located across the vertical  merid- ian in the context of hemianopia, such that  one half of the figure falls within the blind  hemifield. It can occur whether the figure is itself  complete (veridical completion) or incomplete  (paracompletion). Psy- chophysical evidence suggests  that this phenomenon may be a constructive one, and  may share features with completion phenomena in  normal vision. The neural structures mediating  hemianopic completion are unknown. Here we studied  the neural activity evoked by hemianopic completion  using event-related fMRI in an individual (POV) with  a large right visual field homonymous hemianopic  scotoma due to left occipital damage. Either a large  achromatic circular contour straddling the vertical  meridian or a semi- circular contour within the left  hemifield just crossing the vertical meridian was  presented to POV on each trial. POV indicated by  button press whether he perceived a semicircular  contour, a patchy circular contour or a complete  circular contour. On trials where he reported  perceiving a complete circular contour despite being  presented with a semicircular contour  (paracompletion), activity was increased in a region  of ipsilateral extrastriate cortex (contralateral to  the lesion, ipsilateral to the illusory edge of the  circle). These results are discussed in the context  of illusory contour completion in healthy subjects  and more generally in the recovery of function after  brain damage.}  }  @Article{weiskrantz:96,  author = {Lawrence Weiskrantz},  title = {Blindsight revisited},  journal = CurrOpNeuro,  year = 1996,  volume = 6,  pages = {215--220},  abstract = {Some human patients with lesions to their primary  visual (striate) cortex (V1) demonstrate residual  visual capacity, but without acknowledged perceptual  awareness. This phenomenon has been termed  blindsight. Recent results from work on blindsight  patients suggest that it is unlikely to be  attributable to intact residual areas (tags) of  V1. Previous research has reported that blindsight  patients can retain the ability to detect  monochromatic light and grating stimuli, and to  discriminate orientation and direction of movement  in their 'blind' fields. These findings have been  joined by reports that these patients also are  sensitive to, and are able to discriminate,  wavelength in the absence of any experience of  'colour'. This reveals that retinal pathways other  than those to the striate cortex are crucially  involved in vision. Conditions can be controlled for  obtaining either acknowledged awareness or  unawareness of discrimination of the direction of a  small moving target in blindsight patients. This  potentially offers the possibility to determine  whether there are structures uniquely involved in  visual awareness. Monkeys lacking V1 also clearly  demonstrate residual visual capacity, and some  evidence exists that they also experience  'blindsight'.}  }  @Article{winawer.EA:10,  author = {Jonathan Winawer and Hiroshi Horiguchi and Rory  A. Sayres and Kaoru Amano and Brian A. Wandell},  title = {Mapping {hV4} and ventral occipital cortex: The  venous eclipse},  journal = JoVis,  year = 2010,  volume = 10,  number = 5,  pages = {1--22},  abstract = {While the fourth human visual field map (hV4) has  been studied for two decades, there remain  uncertainties about its spatial organization. In  analyzing fMRI measurements designed to resolve  these issues, we discovered a significant problem  that afflicts measurements from ventral occipital  cortex, and particularly measurements near hV4. In  most hemispheres the fMRI hV4 data are contaminated  by artifacts from the transverse sinus (TS). We  created a model of the TS artifact and showed that  the model predicts the locations of anomalous fMRI  responses to simple large-field on–off stimuli. In  many subjects, and particularly the left hemisphere,  the TS artifact masks fMRI responses specifically in  the region of cortex that distinguishes the two main  hV4 models. By selecting subjects with a TS  displaced from the lateral edge of hV4, we were able  to see around the vein. In these subjects, the  visual field coverage extends to the lower meridian,  or nearly so, consistent with a model in which hV4  is located on the ventral surface and responds to  signals throughout the full contralateral  hemifield.}  }  @Article{wu.EA:10,  author = {Zhenghua Wu and Dezhong Yao and Yu Tang and Yilan  Huang and Sheng Su},  title = {Amplitude modulation of steady-state visual evoked  potentials by event-related potentials in a working  memory task},  journal = JBP,  year = 2010,  volume = 36,  pages = {261--271},  abstract = {Previous studies have shown that the amplitude and  phase of the steady-state visual-evoked potential  (SSVEP) can be influenced by a cognitive task, yet  the mechanism of this influence has not been  understood. As the event-related potential (ERP) is  the direct neural electric response to a cognitive  task, studying the relationship between the SSVEP  and ERP would be meaningful in understanding this  underlying mechanism. In this work, the traditional  average method was applied to extract the ERP  directly, following the stimulus of a working memory  task, while a technique named steady-state probe  topography was utilized to estimate the SSVEP under  the simultaneous stimulus of an 8.3-Hz flicker and a  working memory task; a comparison between the ERP  and SSVEP was completed. The results show that the  ERP can modulate the SSVEP amplitude, and for  regions where both SSVEP and ERP are strong, the  modulation depth is large.}  }  @Article{yan.gao:11,  author = {Zheng Yan and Xiaorong Gao},  title = {Functional connectivity analysis of steady-state  visual evoked potentials},  journal = NeuroLett,  year = 2011,  volume = 499,  pages = {199--203},  abstract = {In this paper, functional connectivity of  steady-state visual evoked potential (SSVEP) was  investigated. Directed transfer function (DTF) was  applied to cortical signals recorded from  electroencephalography (EEG) in order to obtain  connectivity patterns. Flow gain was proposed to  assess the role of the specific brain region  involved in the information transmission process. We  found network connections exist in many regions  beyond occipital region. Flow gain mapping both in  8–12 Hz and 13–30 Hz showed that parietal region  seemed to serve as the sole hub of information  transmission. Further studies of flow gain obtained  from channel Pz showed two distinct peaks centered  at about 12 Hz low frequency and 20 Hz medium  frequency respectively. The low frequency region had  a larger value of flow gain. The present study  introduced functional connectivity into  SSVEP. Furthermore, we put forward the concept of  flow gain for the first time to explore the exchange  and processing of brain information during SSVEP.}  }  @Article{zeki.ffytche:98,  author = {S. Zeki and D. H. ffytche},  title = {The {Riddoch} syndrome: insights into the  neurobiology of conscious vision},  journal = {Brain},  year = 1998,  volume = 121,  pages = {25--45},  annote = {We have studied a patient, G.Y., who was rendered  hemianopic following a lesion affecting the primary  visual cortex (area V1), sustained 31 years ago,  with the hope of characterizing his ability to  discriminate visual stimuli presented in his blind  field, both psychophysically and in terms of the  brain activity revealed by imaging methods. Our  results show that (i) there is a correlation between  G.Y.’s capacity to discriminate stimuli presented in  his blind field and his conscious awareness of the  same stimuli and (ii) that G.Y.’s performance on  some tasks is characterized by a marked variability,  both in terms of his awareness for a given level of  discrimination and in his discrimination for a given  level of awareness. The observations on G.Y., and a  comparison of his capacities with those of normal  subjects, leads us to propose a simple model of the  relationship between visual discrimination and  awareness. This supposes that the two independent  capacities are very tightly coupled in normal  subjects (gnosopsia) and that the effect of a V1  lesion is to uncouple them, but only slightly. This  uncoupling leads to two symmetrical departures, on  the one hand to gnosanopsia (awareness without  discrimination) and on the other to agnosopsia  (discrimination without awareness). Our functional  MRI studies show that V5 is always active when  moving stimuli, whether slow or fast, are presented  to his blind field and that the activity in V5  co-varies with less intense activity in other  cortical areas. The difference in cerebral activity  between gnosopsia and agnosopsia is that, in the  latter, the activity in V5 is less intense and lower  statistical thresholds are required to demonstrate  it. Direct comparison of the brain activity during  individual ‘aware’ and ‘unaware’ trials, corrected  for the confounding effects of motion, has also  allowed us, for the first time, to titrate conscious  awareness against brain activity and show that there  is a straightforward relationship between awareness  and activity, both in individual cortical areas, in  this case area V5, and in the reticular activating  system. The imaging evidence, together with the  variability in his levels of awareness and  discrimination, manifested in his capacity to  discriminate consciously on some occasions and  unconsciously on others, leads us to conclude that  agnosopsia, gnosopsia and gnosanopsia are all  manifestations of a single condition which we call  the Riddoch syndrome, in deference to the British  neurologist who, in 1917, first characterized the  major aspect of this disability. We discuss the  significance of these results in relation to  historical views about the organization of the  visual brain.}  }  @Article{zhang.EA:11,  author = {Peng Zhang and Keith Jamison and Stephen Engel and  Bin He and Sheng He},  title = {Binocular Rivalry Requires Visual Attention},  journal = {Neuron},  year = 2011,  volume = 71,  pages = {362--369},  abstract = {An interocular conflict arises when different images  are presented to each eye at the same spatial loca-  tion. The visual system resolves this conflict  through binocular rivalry: observers consciously  perceive spontaneous alternations between the two  images. Visual attention is generally important for  resolving competition between neural  representations. How- ever, given the seemingly  spontaneous and auto- matic nature of binocular  rivalry, the role of attention in resolving  interocular competition remains unclear. Here we  test whether visual attention is necessary to  produce rivalry. Using an EEG frequency-tagging  method to track cortical representations of the con-  flicting images, we show that when attention was  diverted away, rivalry stopped. The EEG data further  suggested that the neural representations of the di-  choptic images combined without attention. Thus,  attention is necessary for dichoptic images to be  engaged in sustained rivalry and may be generally  required for resolving conflicting, potentially  ambig- uous input and giving a single interpretation  access to consciousness.}  }