In 4 SEEG patients of OpenNeuro ds004752 (session 1, first 120-248 s, 6-18 scalp channels), the scalp EEG average at hippocampal IEDs was not larger than a trial-latency-matched null

We tested if interictal epileptiform discharges (IEDs) in hippocampal SEEG contacts give an average deflection on the simultaneous scalp EEG. Data: OpenNeuro ds004752 (verbal Sternberg task), session 1, first 32 MB of each SEEG file (120-248 s). A simple detector (10-60 Hz envelope above 5 x median, bipolar pairs of the deepest hippocampal contacts) found 59-131 IEDs in 4 subjects. Two subjects were excluded (sub-01: 14 IEDs; sub-10: shaft names did not match the pre-registered rule). The statistic was the RMS over scalp channels of the IED-triggered mean (common average reference, 1-30 Hz, -50..+150 ms). The null put each IED at the same within-trial latency in a different trial, which controls for the task ERP. z was above 0 in 2 of 4 subjects (sign test p = 0.69; Fisher p = 0.16). No subject was significant after Holm correction. The only uncorrected effect (sub-06, p = 0.022) had its peak at Fp2, where eye activity is large. Limits: partial data, small and different scalp montages, no visual review of events. The plan was hashed before analysis, and two deviations are listed. The package has numpy-only code, a download script, S3 ETags, local SHA-256 and per-subject averages.

intracranial EEG · scalp EEG · interictal epileptiform discharges · hippocampus · epilepsy · permutation test · verbal working memory