Connectome dtifit bias inhomogeneity correction or not

[andraderenew]

Dear community:

I have found this entry in the mrtrix form and I wonder if it is correct to compute dtifit with the preprocessed dwi data after bias inhomogeneity correction or is it better before that step (without the bias correction).

Beginner: Connectome pipeline (MRtrix3, FSL and Freesurfer)

Hi MRtrix Community, I wanna to study a patient dataset. So, I followed ElijahMak’s workflow to generate the connectome pipeline. However, it is not recommended to segment brain structure by FreeSurfer. So, I tried to use FSL to generate the connectome. My understanding is that I just need to use FSL instead of FreeSurfer in the registration step. I mainly follow their work to do registration by using FLIRT. I would really appreciate it I could get some advice on the basic workflow below. Also…

dtifit -k DTI_30_average-2_denoise_preproc_biascorrected.nii.gz -o DTI_30_average-2_denoise_preproc_biascorrected -m DTI_30_average-2_denoise_preproc_biascorrected_mask.nii -r DTI_30_average-2_denoise_preproc_biascorrected.bvecs -b DTI_30_average-2_denoise_preproc_biascorrected.bvals

Best,
Rene.

[mblesac]

Hi,

If I understnd correctly, dwibiascorrection doesn’t affect to the outputs of dtifit. So you can do it before or after.

Best regards,

Manuel

[bjeurissen]

Note that you can also perform tensor fitting with MRtrix using the dwi2tensor command.