Hi Alfonso,

I’m also working on a longitudinal project with 4 sessions (time points) per subject. Preprocessing completed without any errors as far as I can tell, and functional files (no structural) appear correctly defined in the Setup tab.

However, in the QA plots, many subjects show NaN values, while others display normal-looking results (see attached file). A few subjects are missing 1–2 sessions due to dropouts or no-shows, and for those I only included the available sessions in Setup. But the issue with NaN values also appears in subjects who have all 4 sessions complete and seemingly correctly processed.

So far, I’ve checked the detup tab: all functional images are defined correctly and log file: no obvious errors or warnings.

I’m not sure if the missing sessions in some subjects are somehow affecting QC, or whether there's another reason that subjects with apparently complete data would show up as NaN.

Also, I wanted to ask if you have any best practice advice for working with multiple-session data, particularly when some sessions may have significantly more motion than others. Is there a recommended way to check session-wise QA, and how would you suggest dealing with that situation?

Thanks in advance!

Best,

Asena