Hello Clinica team,
I am using the adni-to-bids pipeline for T1 files and I am unsure about what files are selected by the converter.
Indeed the
documentation specifies that when preprocessed files are available, the converter will look for the file that has been "Scaled" (i.e. fully preprocessed) and then use the "N3" image (the preprocessed image not scaled). This explains why the "Image_ID" column in t1_paths.tsv can refer to the scaled image while the "Sequence" column refers to the N3-image (as explained in the documentation). Explicitly wanting to select the N3-image (not scaled) has been confirmed in this
GitHub Issue, and this
post on the forum. So I would expect the converter to select the image based on the information in the "Sequence" column to select the N3-image.
However the
_find_path_single_image function uses the "Image_ID" column to select the image files (
in this line). Therefore it seems that the scaled image, to which the image id refers, is actually converted to BIDS instead of the N3-image.
Did I understand the code correctly? And do you confirm that the N3 image should be used instead of the scaled image?
If so, then I think the _find_path_single_image function should use the "Sequence" column to select the file rather than "Image_ID".
Thanks,
Best regards,
Pauline