About Multishell dMRI processing

1

Hi all,

My dMRI dataset is a multi-shell acquisition that includes b-values of 0, 800, and 2000.

I used the following command to estimate the response functions:

dwi2response tournier $datadir/${a}/${a}_dwi_den_deringed_preproc_unbiased_normalized.mif $datadir/${a}/${a}_response.txt -scratch $datadir/${a}/

I then used the following command to compute the WM FODs:

dwi2fod msmt_csd $datadir/${a}/${a}_dwi_den_deringed_preproc_unbiased_normalized_upsampled.mif $datadir/group_average_response.txt $datadir/${a}/${a}_wmfod.mif -mask $datadir/${a}/${a}_dwi_mask_upsampled.mif -shells 2000 -force

My question is: does this code use all the shells in my dMRI data?
If not, which part of the code should I modify?

Thanks in advance,

2

Hi @armin,

I think the tournier algorithm doesn’t use all the multi-shell data, for that, you need to use dhollander.

Then in the dwi2fod call, you need to add the three tissue responses and remove the -shells flag.

Best regards,

Manuel