Hello everyone,
i have a question regarding event/task-related EEG data. Are there any indications that baseline power increases from trial to trial? For example, in the case of a fine motor task (e.g. finger movement), that baseline power slowly accumulates from trial to trial through the task (baseline power trial 1 < trial 100)?
I am aware that the time between trials should be such that there is a return to baseline. However, if the power increases slowly, could it be that an effect is only seen after 70-80 trials?
Does anyone have experience or know of studies that can provide guidance on this issue?
Regards
Niko
Hi, Nikolas Rizzi
Given that the EEG signal is non-stationary, I'd say that it is perfectly possible for the signal to 'drift' across trials. In fact, I have at some point recorded data with such a behaviour. In any case, there are drifts that can be removed with the appropriate type of filtering (see, for example, this link: https://benediktehinger.de/blog/science/electrode-drift-in-eeg/). I would recommend normalizing your task data by the power at baseline, so that, if such a drift is indeed present, you can consider it in both task and baseline.
You could also always compute the baseline power and check if it, indeed, increases over time between your trials.
Does that make sense?
The power changes over time, which is unavoidable. You can try to alleviate that by taking into consideration only the average of the baselines in the last N trials. Alternatively, you can apply Stationarity Subspace Analysis, which I used in my project and is described here:Article Effect of power feature covariance shift on BCI spatial-filt...
The more key question is whether the nonstationarity should be considered undesirable, as can be also caused by neural activity and not only by the measuring equipment?!
- Badges
- Science topic