It’s still not 100% clear exactly what you’re trying to achieve, but I’ll throw in a couple of possibilities:
-
If you register each subject to a template, and define your rectangular ROI in template space, you could then transform that back to each individual subject. Whether or not this is appropriate depends on the particulars of your experiment that are still slightly unclear.
-
While the
mreditcommand does not currently have the capability to draw rectangular cuboids, this would not be a very difficult capability to implement. It would however have the constraint that the shape would be axis-aligned, rather than being an XYZ rectangular cuboid relative to the scanner axes re-sampled onto a particular voxel grid.
Rob