Morphometry conventions
effigies opened this issue · comments
As discussed with DCAN lab today, we will continue using GIFTI morphometry images following FreeSurfer conventions. These depart from HCP conventions by a sign flip in the cases of curvature and sulcal depth. Thickness is an absolute value, and so there is no difference.
For CIFTI-2 .dscalar.nii files, we will adopt the HCP conventions, as the only tooling that exists to work with these adopts these.
This will be an unfortunate inconsistency in our outputs, but it was felt that adopting either convention wholesale would produce more confusion. This will need to be stated clearly in the documentation. We may want to add a desc-<FS|HCP> to indicate the convention in each case.
Done in nipreps/fmriprep#3137