Baseline selection of Epileptogenicity map

Discussions
,
1

I would like to use the Epileptogenicity Map to localize the symptom-onset zone. The issue I’m facing is that the symptom I want to study always appears during the seizure rather than as the initial manifestation. By that time, the abnormal EEG activity has already spread to many electrode contacts. In the Brainstorm tutorial, the baseline is chosen from EEG recordings prior to seizure onset. Could I instead use the EEG segment just before the appearance of this particular symptom as the baseline? If so, what would be the optimal length of time in seconds?

2

Hello,

I would like to use the Epileptogenicity Map to localize the symptom-onset zone. The issue I’m facing is that the symptom I want to study always appears during the seizure rather than as the initial manifestation. By that time, the abnormal EEG activity has already spread to many electrode contacts.

Is it possible to use the EEG segment just before the symptom onset as the baseline, rather than using a clean interictal recording without discharges—even if at that point the abnormal activity has already spread throughout the brain?

Thank you very much for your guidance!

3

Is this the same post as in?

@yashvakilna, any suggestion on this subject?

4

This post appears to have been made by my colleague—we have the same question

5

Both baseline options are valid. Please choose the one that aligns with your null hypothesis, and ensure the baseline segment is sufficiently stationary to yield a reliable noise covariance estimate.

Choice guided by the null hypothesis:

  • Null hypothesis: “Activity at the time of interest equals background noise.”
    Use a clean pre-ictal segment before seizure onset, sufficiently separated from the analysis interval. This asks the question: whether the current activity departs from the resting background.
  • Null hypothesis: “Activity at the time of interest equals the immediate pre-symptom state.”
    Use the segment immediately preceding the symptom's appearance. This asks the question: whether a change specifically accompanies symptom emergence, even if the brain is already ictal.

Duration:

  • The tutorial’s 5 to 30 seconds is sensible. Prioritize stationarity over length. If the baseline drifts or ramps up, several shorter clean segments are preferable to one long contaminated segment.