Use quantitative T1 map when available

[tsalo]

Summary

In the two-compartment model that we use for multi-delay data, we need an estimate for the T1 of gray matter. At the moment, we use a single value across the brain; however, it would be nice to use quantitative T1 maps when possible.

Additional details

This should improve CBF estimates.

However, it may be difficult to collect the T1 maps, given that quantitative maps like T1map should generally be organized in derivative datasets, rather than raw ones. I wonder if anyone has developed any tricks for feeding derivatives into BIDS Apps?

Next steps

1. Load derivatives dataset, when available.
2. Query the raw and derivatives datasets for T1map files associated with the same subject.
3. Determine space of the T1map file.
4. Find set of transforms from T1map's space to the ASL reference space.
5. Warp the T1map to the ASL reference space.
6. Apply the GM mask to the T1map?
7. Use the GM T1 values in voxel-wise CBF estimation.
   • What about WM?

[tsalo]

We'll be acquiring T1maps with our ASL scans in a new study, so I need to get this working at some point.

[tsalo]

Starting to think about this. I might place the T1maps in the raw dataset for this initial attempt, and then work on generalizing the derivative collection later.

[tsalo]

Use BBRegister command from sMRIPrep: https://github.com/nipreps/smriprep/blob/b839befd008e02f329c59aa634583a18e4049829/src/smriprep/workflows/anatomical.py#L1164-L1173

Consider generating mask in T1 map calculation pipeline.

[tsalo]

We can't use the T1 map for CBF/ATT estimation with single-PLD data, so I'm downgrading this on the priority list.