Dear Neurostars,
I am trying on trying out some methods to improve the attached coregistration. Mostly, I noticed the overlay around the anterior temporal lobe, rostral anterior cingulate, and cerebellum aren’t great. So, I want to compare the fmriprep coregistration (using mri_coreg/freesurfer BBR) with ANT. To my understanding, the fmriprep coregistration step uses this as the moving image: “sub-*/func/sub-*_task-*_desc-hmc_boldref.nii.gz”, and this as the fixed image: “freesurfer/sub-*/mri/brainmask.mgz”. Am I correct?
If so, would the following steps using ANTS be comparable to the fmriprep coregistration (assuming I want to keep the same resolution for the moving image like fmriprep does by default)?
mri_convert freesurfer/sub-*/mri/brainmask.mgz freesurfer/sub-*/mri/brainmask.nii.gz
##resampling the fixed image to native space
ResampleImage 3 freesurfer/sub-*/mri/brainmask freesurfer/sub-*/mri/brainmask_res2.5.nii.gz 2.5x2.5x2.5 0 1
##resample moving image to above image
antsApplyTransforms -i sub-*/func/sub-*_task-*_desc-hmc_boldref.nii.gz freesurfer/sub-*/mri/brainmask_res2.5.nii.gz -o $fprepdir"/func/testANTs.nii.gz" -n NearestNeighbor
Also, does the figure in the html output (figures/sub-desc-coreg_bold.svg) shows the func/subdesc-coreg_boldref.nii.gz file as the moving figure and the brain mask in the subject’s freesurfer directory as the fixed image?
Thank you for your help.
Best,
Kareem Al-Khalil
