Dear Community, 

In my lab we have recently developed a high-performing pipeline able to automatic reconstruct the anterior optic pathway (that is cool)

Unfortunately, we are experiencing  big troubles registering the morphological T1 images to the b0 image, making a DICE comparison impossible for quality check

Does anyone know a good non-linear registration tool able to efficiently register the b0 image to the T1 ? 

That would be so lovely

Thank you in advances 

Laura Ludovica Gramegna !

Check out the tools from the TORTOISE group:

https://tortoise.nibib.nih.gov/software_...

Anthony

Hi Laura,
The registration errors are likely to result from distortions in the B0 images. Are you using any tools to correct for distortion? FSL had some, for example, that required an estimate of the field map (per-subject B0 measurement) but would probably get you closer with even a generic field map - using an average one for all of your subjects, since the distortions will be somewhat similar.
Otherwise you have a challenge, both distorted DTI source images and differences in image contrast between the B0 and the T1. Do you have a T2w image for your subjects?
Regards,
Alex