Poor registration results when coresistering T1w and MT data

[ryu]

Dear SCT Team:
The coregistration process between MTw and T1w images has been unsuccessful. Could you please suggest methods to improve the registration outcome?
This is the data:
invivo_MTw.nii.gz (1.1 MB)
invivo_MTw_seg.nii.gz (2.5 KB)
invivo_T1w.nii.gz (1.1 MB)
invivo_T1w_seg.nii.gz (2.5 KB)
mask_invivo_T1w.nii.gz (15.6 KB)

[joshuacwnewton]

Dear @ryu,

Thank you for your question, and for sharing the data.

Could you please also share the commands that you have tried so far? (e.g. Which segmentation commands, which labeling commands, registration commands, etc.)

Also, could you please show the unsuccessful registration results? What went wrong?

Thank you kindly,
Joshua

[ryu]

The command I used is as follows：
sct_register_multimodal -i invivo_MTw.nii -iseg invivo_MTw_seg.nii -d invivo_T1w.nii -dseg invivo_T1w.nii -m mask_invivo_T1w.nii -param step=1,type=seg,algo=centermass:step=2,type=seg,metric=CC,algo=bsplinesyn,slicewise=1,iter=3:step=3,type=im,metric=CC,algo=syn,slicewise=1,iter=2 -qc qc

[joshuacwnewton]

Thank you for sharing the command! I am able to reproduce the faulty registration.

@ryu

sct_register_multimodal -i invivo_MTw.nii.gz
                        -iseg invivo_MTw_seg.nii.gz
                        -d invivo_T1w.nii.gz
                        -dseg invivo_T1w.nii.gz
                        -m mask_invivo_T1w.nii.gz 
                        -param step=1,type=seg,algo=centermass:step=2,type=seg,metric=CC,algo=bsplinesyn,slicewise=1,iter=3:step=3,type=im,metric=CC,algo=syn,slicewise=1,iter=2 
                        -qc qc

I notice that the invivo_T1w.nii.gz file is used for both -d and -dseg. Switching -dseg to be invivo_T1w_seg.nii.gz fixes the registration:

Kind regards,
Joshua