Compatibility issue with XCP-D requiring T1 when using T2-only input in NiBabies 24.0.1

[yanbin-niu]

Hi NIBABIES team,

We’re currently using NiBabies v24.0.1 to process newborn MRI data. One really helpful feature is the ability to run the pipeline with either T1- or T2-weighted images, which is especially important for newborns where T2 contrast is often more informative. We always prioritize T2 acquisitions over T1 in our newborn protocol, and we ended up having a bunch of data with only T2 (not T1).

However, we’ve run into a downstream issue when running XCP-D on the outputs. It seems that XCP-D expects a T1 image to be present and fails when only a T2 image is available.

I’m unsure whether this is best reported here or directly to XCP-D, but I wanted to raise it as it seems like a compatibility issue between the two tools. I checked the recent release notes and didn’t see anything addressing this.

Would appreciate any guidance or thoughts! for exmple, should this be reported to XCP-D as a feature request for T2-only support?
Is there a workaround or recommended approach when using NiBabies with T2-only inputs? (I am currently thinking to create a peudo t1 from t2?)

Thanks for your help, and appreciate all the work going into these tools!