Mouse DTI data


I am new to mouse DTI data, and was wondering whether standard preprocessing (denoising, unringing, bias field and distortion correction etc…) and DTI measures’ extraction + tractography for human data is applicable also to rodent data, or if ad-hoc changes (and in case which ones) have to be made.

thanks in advance,

All of the approaches are broadly applicable, but is dependent on the type of data you have as to whether various pipeline pieces are directly translatable.

The one semi-major caveat is bias correction, as the grid size, iterations, and smooth factors are preset for human brains and need to be better optimize for your image dimensions and voxel size.

1 Like

Thank you very much. Once I have the data, I’ll try to implement default Mrtrix3 preprocessing pipeline according to their specific appearance.

Be very wary of any command that relies on accurate brain masking. Many scripted commands will use dwi2mask if no mask is supplied and the resulting mask might be very off. Best to create the brain mask yourself and check it visually and pass it as an option.

1 Like

Thank you very much. So do you suggest create a mask manually slice per slice?

Moreover, what about 5TT segmentation and use of a proper atlas to create parcellation image for tractography+connectomics?