In 10 patients, movement related SEEG high gamma similarity fell to near zero beyond 20 mm in template coordinates

We measured how similarity in movement related 70–150 Hz high gamma activity (HGA) changes with distance between bipolar SEEG channels in 10 patients (844 positioned channels) performing a mouse movement task. We used archived OpenNeuro ds006253 window features and fsaverage electrode files previously prepared in SCI contributions. For each channel, we calculated a standardized log10 HGA change from rest to 0–0.5 s after movement onset on odd and even trials separately, then correlated channel effects across the two trial halves by distance; pairs sharing a contact were excluded. Mean pair correlation was 0.23 at 0–10 mm, 0.05 at 10–20 mm, 0.01 at 20–30 mm, and approximately zero beyond 30 mm. Same channel split half reliability was 0.76. Pre registered sign flip tests against the 50+ mm bin were significant at 0–10 mm (9/10 patients, p=0.0049), 10–20 mm (8/10, Holm p=0.033), and 20–30 mm (8/10, Holm p=0.040); the last contrast was driven mostly by the negative 50+ mm value. Exploratory finer bins showed normalized similarity of 0.25 at 5–10 mm, 0.17 at 10–15 mm, and 0.01 at 15–20 mm. Same shaft and different shaft pairs within 20 mm did not differ significantly (p=0.23), and effect magnitude did not cluster in space (p=0.28). The planned half length could not be estimated because normalized similarity was already 0.30 in the first 0–10 mm bin; its planned bootstrap confidence interval was invalid. These findings suggest that movement related HGA effects are similar over short distances and approximately independent beyond about 20 mm, within the limits of fsaverage template distances, one task and time window, bipolar referencing, and 10 patients.

SEEG · high gamma activity · movement · electrode distance · intracranial EEG · brain computer interface · split half reliability