We used Mindboggle (
http://mindboggle.info), an open source brain morphometry platform, to obtain measures of volume, surface, and tabular data containing labels, features, and shape formation of the brain surface. A complete description of Mindboggle's algorithms can be found in Kliein
\cite{Klein_2016} or in the software documenation at
http://mindboggle.info/software.html.
Briefly, the inputs are the outer cortical surfaces meshes generated by Freesurfer 6.0 as described in the previous section. To refine segmentation, labeling and volume shape analyses, Mindboggle also takes output from Advanced Normalization Tools (ANTs) (website download:
https://github.com/stnava/ANTs.git) which performs brain volume extraction, tissue-class segmentation, and registration-based labeling. To generate the ANTs files used by Mindboggle, we ran the following script. Note that the template used was OASIS-30 Atropos template downloaded from
https://osf.io/rh9km/download. The -n argument is the number of iterations of n4biasfieldcorrection on probability estimates for each of the tissues segments, and the -x argument is the number of iterations to determine cortical thickness. The -q argument indicates that antsRegistrationSyNQuick.sh will be the basis for registration during brain extraction.