Issue with template warping onto chunk 3 of dwi images

[Kaonashi22]

The template warping onto chunk 3 of dwi images partially failed partially (on the last slices, including T5) or completely for a few subjects,. The diffusion metrics are not extracted at these levels. Here are screenshots for two subjects:

I can share the images of these subjects

[jcohenadad]

The issue is caused by the way the DWI data were acquired. As you can see below for sub-DEV208Sujet11, the uppermost slice of chunk3 is at T4 vertebral level, whereas the T2w anatomical scan covers down to the mid-T4 level:

Given that the code is designed to extract metrics between C2 and T5, it is not surprising that this subject (and other subjects with inappropriate FOV placement) is problematic:

https://github.com/sct-pipeline/spine-park/blob/924e332c3b4836baa087ea740a7837120d0b7cbf/batch_processing.sh#L20

[!Note] Also note the important gap between chunk1 and chunk2

Because we are now registering the data using all visible disc labels (https://github.com/sct-pipeline/spine-park/pull/49), any data acquired below the bottom visible disc (as in the case for chunk3 here) won't be registered to the anatomical data.

[Kaonashi22]

Ok, thanks @jcohenadad; then I will have to exclude these subjects from the dwi analysis.