Hi Raymundo/Brainstorm team,
I imported an MRI into Brainstorm after running it through FreeSurfer. We have stroke patients that have often have a damaged cortex, so the registration between cortex and MRI is always something to look at, but I have not seen anything like this registration yet (screenshot below).
How concerned should I be about this cortical registration to the MRI? This "lack of accuracy" is around where the lesion is.
Thank you,
Adam
@Anand_Joshi can you please check this.
@bakerdon Most of the standard MRI processing tools (freesurfer, SPM, Brainsuite, ...) do not support segmenting lesioned brains.
In the image you shared, the voxel intensity around the lesion is not detected as brain tissue, which is a common issue in this case.
You may need to edit manually the outputs, depending on what you want to do next.
Some new methods have recently been developed using deep learning that allows better segmentation of such MRIs. Still, I'm unaware if any are already distributed as software for the community.
