Investigate failed denoising subjects

NBCLab / power-replication

A replication and extension of Power et al. (2018)

https://www.overleaf.com/read/swgjxcjqytxg

Apache License 2.0

2 stars 0 forks source link

Investigate failed denoising subjects #15

Closed tsalo closed 2 years ago

tsalo commented 2 years ago

Dataset dset-camcan subject sub-CC110187 missing MEDN file.
Dataset dset-camcan subject sub-CC110411 missing MEDN file.
Dataset dset-camcan subject sub-CC310142 missing MEDN file.
Dataset dset-camcan subject sub-CC420587 missing MEDN file.
Dataset dset-camcan subject sub-CC620026 missing MEDN file.
Dataset dset-dupre subject sub-16 missing:
- sub-16_task-rest_run-1_desc-RVReg_bold.nii.gz
- sub-16_task-rest_run-1_desc-RVTReg_bold.nii.gz

tsalo commented 2 years ago

The CamCAN failures seem to be due to failed SVDs during the tedana workflow.

tsalo commented 2 years ago

sub-16 failed due to missing physio regressors. Traceback when running the peakdet step:

Traceback (most recent call last):
  File "05a_generate_physio_confounds.py", line 737, in <module>
    _main()
  File "05a_generate_physio_confounds.py", line 732, in _main
    main(project_dir=project_dir)
  File "05a_generate_physio_confounds.py", line 690, in main
    subject,
  File "05a_generate_physio_confounds.py", line 221, in compile_physio_regressors
    assert resp_data.shape[0] >= resp_data_end
AssertionError

tsalo commented 2 years ago

sub-16 doesn't have complete physio data, which explains why it was failing.