I am currently attempting to combine tractography results (streamlines) from around 100 different young healthy normal subjects into a common space (e.g. MNI). I have 56 direction HARDI images. I am wondering if anyone has input as to what is the best way to do this. I have narrowed the process down to one of the following processes:
1) First construct the ODF's in diffusion image space and then coregister them to MNI space and perform tractography. However, I have not been able to easily find ODF coregistration tools in any of the common diffusion software packages (am I missing something?), but I have read a few papers that do this. Does anyone know of a good ODF co-registration tool?
2) First linearly coregister the b0 image to an MNI-EPI template (using FSL, for example) and apply this transformation to each diffusion image as well as the b-vector. Then I would reconstruct the ODF in MNI space and perform tractography. The obvious disadvantage of this method is that you are performing the coregistration to common space using only a linear transformation. Are there other problems in doing this?
3) First construct the ODF and perform tractography in diffusion space, then apply a non-linear transformation to each point on the streamline to obtain the corresponding streamline in MNI space. I have found that dipy has a code to do this - anyone know of another software that can do this?
Thanks so much for any help you can offer!