Dear Dr. Yeh,
I am working on a project that involves resampling Julich probabilistic ROIs (originally in MNI space) to native diffusion space for analysis and tractography in DSI Studio. My current pipeline uses ANTs to first transform ROIs from MNI to native T1 (ACPC) space, then resamples these masks into native DWI space (using the preprocessed DWI from QSIPrep as the reference image). I apply a 0.3 threshold to binarize the probabilistic masks before resampling.
While this approach appears to work, I am unsure if there are best practices or specific recommendations to ensure optimal alignment and compatibility with DSI Studio. Additionally, I want to confirm that resampling using the DWI reference image aligns the masks to the b=0 volume, and that this approach is appropriate for use with DSI Studio’s segmentation and tractography workflows.
Do you have any advice or suggestions for improving this workflow, or common pitfalls I should watch out for when preparing ROI masks for DSI Studio?
Thank you very much for your time.
Sincerely,
Yaojie