--skip_frame parameter keeps skipped frames in output file

[BoletuPeng]

Environment

• DeepPrep version: 25.1.0

Command

(deepprep) root@03a390c4ea00:~# bash /opt/DeepPrep/deepprep/web/pages/postprocess.sh \
  /output/BOLD /output/postproc participant \
  --fs_license_file /fs_license.txt \
  --confounds_index_file /confoundlist.txt \
  --task_id rest \
  --space 'MNI152NLin6Asym fsaverage6' \
  --skip_frame 5 \
  --surface_fwhm 0 \
  --volume_fwhm 0 \
  --bandpass '0.001-9999'
INFO: args: /output/BOLD /output/postproc participant --fs_license_file /fs_license.txt --confounds_index_file /confoundlist.txt --task_id rest --space MNI152NLin6Asym fsaverage6 --skip_frame 5 --surface_fwhm 0 --volume_fwhm 0 --bandpass 0.001-9999
Input --fs_license_file : /fs_license.txt
FREESURFER_HOME is set to: /opt/freesurfer
INFO: existed nextflow WorkDir: /output/postproc/WorkDir/nextflow
INFO: run_config : /output/postproc/WorkDir/nextflow/run.config
.......................

Observed behavior

• Input: 242 timepoints
• Output (desc-regression_bold.nii.gz): still 242 timepoints
• First 5 frames contain real data (non-zero, similar statistics to other frames)

Expected behavior

Either:

1. Output should have 237 timepoints (skip_frame=5 removed), or
2. Documentation should clarify that --skip_frame only excludes frames from regression calculation but retains them in output

Question

Were the first 5 frames excluded from regression model fitting? If so, should they be removed from output or marked somehow?