Registration of AAL atlas to subjects' T1

1

Hello all,

I want to use the HCP data to construct connectome of AAL atlas, I have registered subject’s T1 to DWI, then using flirt and fnirt to register the AAL atlas in MNI space to subject’s T1 space to serve in tck2connectome, but the output AAL2T1 was weird, could anyone help me?

flirt -ref meanb0.nii.gz -in ../raw_data/T1w_acpc_dc_restore_brain.nii.gz -omat T12DWI_affine.mat
flirt -ref ${FSLDIR}/data/standard/MNI152_T1_1mm_brain -in ../raw_data/T1w_acpc_dc_restore_brain.nii.gz -omat T12MNI_affine.mat
convert_xfm -omat MNI2T1_affine.mat -inverse T12MNI_affine.mat
convert_xfm -omat MNI2DWI_affine.mat -concat T12DWI_affine.mat MNI2T1_affine.mat
fnirt --ref=../raw_data/T1w_acpc_dc_restore_brain.nii.gz --aff=MNI2DWI_affine.mat --in=${FSLDIR}/data/standard/MNI152_T1_1mm_brain --cout=MNI2DWI_nonliear
applywarp --ref=meanb0.nii --in=/Users/bclab/Desktop/mask/aal.nii --warp=MNI2DWI_nonliear --out=aal2DWI.nii.gz

image
image.jpg1071×1268 124 KB

2

Hi all,
I have solved my problem by using the nearestneighbor interpolation method of Flirt as below, for people who have same questions:

image
image.jpg1154×302 173 KB

1 Like
3

I tried to use FSL to segment the subcortical regions of the brain, but there were only 15. How to use Altas to segment hundreds of regions? I tried FreeSurfer, but it took too long. Because I need the command line to complete the operation, Cat12 and SPM can not be completed, at least I did not find. Could you help me?

4

Dear,

I don’t know of any easily available software that can perform sub-cortical segmentation from T1w images apart from FSL FIRST and FreeSurfer. You might find it easier to use the methods described above to move atlas labels to your participant’s diffusion space, at the cost of some specificity.

All the best!