mne-bids-pipeline icon indicating copy to clipboard operation
mne-bids-pipeline copied to clipboard

select and reject ICA components

Open SophieHerbst opened this issue 5 years ago • 4 comments

Right now 04-artifact_correction_ica.py just runs the ICA but doesn't do anything with it. We need a new scripts to visualize and select the components: https://mne-tools.github.io/stable/auto_tutorials/plot_artifacts_correction_ica.html

and to reject them

ica.apply(realData, exclude=exclude_ICA) 

SophieHerbst avatar Jan 31 '19 14:01 SophieHerbst

@agramfort: can we look at how you would select ICA components when we meet next Tuesday? I can make a suggestion, but I am not fully satisfied with how I am doing it right now.

SophieHerbst avatar Feb 13 '19 20:02 SophieHerbst

ok or ping me on hangout

agramfort avatar Feb 14 '19 08:02 agramfort

Hi again! I am putting together an ICA-rejection script (I'll need it for my student). @agramfort – is there any way to run create_ecg_epochs on epochs data? Otherwise I will have to load the raws from all runs again to do this, which takes a lot of memory.

SophieHerbst avatar Feb 17 '19 19:02 SophieHerbst

no as ecg epochs are based on ecg peaks as events. I would recommend to use only one maxfiltered run. It should be more than enough to have a good number of ecg epochs. The alternative is to use epochs but to keep the ECG channel in the epochs and detect ECG ICA sources by correlation which may work. It would require some work and would take a weeks or more to do. I would go with one raw object.

agramfort avatar Feb 17 '19 20:02 agramfort

Closing, automated artifact rejection is working now and ICs selected for rejection can be reviewed and adjusted manually.

hoechenberger avatar Mar 07 '24 10:03 hoechenberger

haha, that was a really old one!

SophieHerbst avatar Mar 07 '24 11:03 SophieHerbst

Yes, at least after 5 years we could finally close it! It's almost done with pre-school :)

hoechenberger avatar Mar 07 '24 11:03 hoechenberger