The PD population under study was broken up into two groups by stimulation.
MRI images used for analysis were captured with the following settings:
Volume analyses were run on a T1-Weighted 3D volumetric scan (1x1x1mm3): 3000 Repetition time (TR) (ms), 90 Echo time (TE) (ms), 90 Inversion time (TI) (ms), 320x178 Matrix, 240x180 Field of View (mm), 64.5 Bandwidth (Hz/Px), axial orientation, 32x2 Slices, and 6 min 48 sec Acquisition time.
NIFIT format
Raw DICOMS were converted to NIFTI format using Statistical Parametric Mapping (SPM 12) in Matlab. NIFTI is a standardized format for analyzing 4D neuroimaging data.
Three separate analyses were conducted using the following datasets:
Segmented grey and white matter of each NIFTI dataset using CAT 12 Voxel-based morphometry (Figure 1). CAT 12 is a neuroimaging technique that measures differences in brain anatomy using statistical analysis (Ashburner and Friston, 2005). Brain area specific region of interest segmentations were extracted using FreeSurfer (Dale et al., 1999; Fischl et al., 1999). This process was used to generate brain area specific estimations of volume. Using NIFTI files, I manually measured bilateral olfactory bulb volumes using ITK-SNAP in the coronal view seen in Figure 2B (Yushkevich et al., 2006). Both left and right OB volumes were traced to allow for an interhemispheric control, as well as compare ipsilateral bulbs (bulb on the hemisphere that received the stimulation).
Statistical Analysis and data organization
For group analyses, we used a Kruskal-Wallis non-parametric ANOVA to compare volumes between PD stimulation groups. Significant group differences were further tested with multiple post-hoc comparisons using a Dunn’s test. To compare between PD and ET we used a non-parametric t-test.