Hello! We are running into some issues when trying to properly preprocess our rodent DTI data. I've laid out the steps we've completed and the stage where we are running into trouble— essentially we are having trouble correcting phase-encoding distortion. Any advice or tips on this would be appreciated!!

*All images were acquired by a small-animal Bruker 9.4T MRI system running Paravision 6.0.1 with a hydrogen receive only surface array coil shaped for rat brains.*

Preprocessing steps we've completed:

1. Converted .dicom to .nifti

2. To fix the phase-encoding distortion (caused by AP or PA sequences, aka blip-up/-down sequences), we first extracted the b0 images from the diffusion scan

3. Then, we manually drew brain vs. non-brain boundary in T2-weighted scan for brain extraction

Challenge: We don’t have a blip-down (PA) sequence, so we aren’t able to use topup for correcting phase-encoding distortion Therefore, we need to utilize a structural scan (in our case, a T2-weighted image) to create an undistorted synthesized b0 image, using Synb0-Disco

Problem: When we attempted to feed rat image into Synb0-Disco, it promoted a lot of error messages, which could be due to misalignment with registered brain atlas (Synb0-Disco was trained using human brains)

Troubleshooting questions:

1. Have animal researcher(s) attempted to create undistorted synthesized b0 images, if so, how do they do it?

2. Alternatively, is it okay to not correct for distortion in animal scans?

Next step after correcting for phase-encoding distortion:

4. Run the data through eddy (FSL) to conclude preprocessing steps Optional: up-sampling the diffusion scan to isotopic voxel space Current voxel size is: 1 x 1 x 7 mm We could make it: 1 x 1 x 1 mm, or 1.5 x 1.5 x 1.5 mm

5. Create a population-based atlas (a template generated by our data) in DSI Studio

6. Register all the individual scans to that population-based atlas