Reporting the data import for manuscript (Remove DC offset => All recordings)

Discussions
1

Hi,
I am a new brainstorm user, writing my first EEG analysis draft. In the method part, I want to explain that I epoched 20 seconds of data useing the option:
Remove DC Offset: select baseline definition
I checked the option:
All recordings: Baseline computed for each output file (since my main baseline is another condition that I epoch separately). However, I can not find this option's characteristic in the tutorial (https://neuroimage.usc.edu/brainstorm/Tutorials/Epoching) to report in my method section. Please kindly let me know the details of how this function work. If removing DC offset is like using high-pass filter with a very low frequency (for instance 0.3Hz), please tell me the filter's features, like its slope and its cut-off frequency. I and probably every other one who is using brainstrom need this info for documentation. Thank you very much for your outstanding support.

2

The algorithm involved when using the option "Remove DC offset" described in this section:
https://neuroimage.usc.edu/brainstorm/Tutorials/Epoching#Import_in_database

When selecting "All recordings", it simply selects all the epoch as the time window to compute the "baseline mean" instead of a subset of the epoch.

If you are using a high-pass filter or band-bass filter on the continuous file before epoching, you probably don't need this option, as the baseline of the signal is already removed (the is no "DC offset" left).
Before processing your own EEG recordings, I recommend you follow the tutorial "EEG and epilepsy":
https://neuroimage.usc.edu/brainstorm/Tutorials/Epilepsy#Band-pass_filter

since my main baseline is another condition that I epoch separately

Here we are only discussing of the arbitrary "DC offset" you might have in your channel at the time of the epoch. This is related with the acquisition hardware more that with the brain activity.
The "brain baseline" (resting period or reference condition), is not to be used here. It will be useful later for normalizing source maps, or contrasting an active state with a rest condition.

1 Like