Question
Question Posted 08/02/20:
Dear ADNI expert,
I have a couple of scans for the same subjects; one is an MRI T1w, and the other is a DTI scan. I noticed that the T1 scan is absolutely closer to the MNI space alignment way much better than the DTI scan although both were scanned during the same session.
Should not they both be closer to each other whatever the distance and alginment with the MNI? Would you please clarify?
Thanks a lot,
Best
Dear ADNI expert,
I have a couple of scans for the same subjects; one is an MRI T1w, and the other is a DTI scan. I noticed that the T1 scan is absolutely closer to the MNI space alignment way much better than the DTI scan although both were scanned during the same session.
Should not they both be closer to each other whatever the distance and alginment with the MNI? Would you please clarify?
Thanks a lot,
Best
Response posted 08/02/20 by Rob Reid:
Hello, I think you are seeing Echo Planar Imaging (EPI) distortion in the DTI. In ADNI (and most other in vivo studies) both dMRI and fMRI are scanned with EPI because it is much faster than the alternatives. However, it has the drawback of being stretched or compressed in the phase encoding direction (y) wherever there are changes in magnetic susceptibility, e.g. air/tissue/bone interfaces. There are multiple ways to correct this, but for ADNI data if EPI/T1w alignment is needed I recommend nonlinearly warping the EPI data to the T1w scan. Do *not* warp the FA to the T1w - instead use the raw diffusion data to construct a T1w simulation, and warp that to the T1w, a la Bhushan, C. et al. Co-registration and distortion correction of diffusion and anatomical images based on inverse contrast normalization. NeuroImage doi:10.1016/j.neuroimage.2015.03.050.
