I have 5 NII files (EPI) for each subject. Initially, I run fmriprep with no problems (i.e. I got 5 confounds.tsv files). However, after looking at the HTML, I have spotted a huge spike (FD>1) in the last 5 volumes of 2 of the files. This is technically not a problem since these 5 volumes are not part of the stimuli, so I thought I could leave out the last 5 volumes using fslroi.
I cut away the last 5 volumes and re-run fmriprep with the modified files. To my surprise, this time I only got 3 confounds.tsv files (for the unmodified files) and a set of 4 files for the modified NII. This set has MELODIC_mixing.tsv, AROMAnoiseICs.csv, scanner_to-T1w_mode-image_xfm.txt, and T1w_to-scanner_mode-image_xfm.txt. MELODIC_mixing.tsv, AROMAnoiseICs.csv do not have the confounds and I am bewildered on why I did not get the confounds.tsv, so any help is highly appreciated!