Distance-only models overestimated hippocampal and amygdala N1 responses in 4 of 6 patients, with no consistent group effect

This study tests if hippocampus and amygdala (MTL) SEEG contacts get early CCEP (N1) responses at the same rate as neocortical gray-matter contacts at equal distance. Data: OpenNeuro ds005448, 8 SEEG patients, expert-checked N1 labels, 18,915 observations at 10-80 mm. A logistic model of log distance and same shaft was fit on neocortical receivers. In 4 of 6 patients with MTL contacts, MTL receivers had fewer N1 responses than the model predicted (median observed/expected 0.49, range 0.00-1.70, pooled 0.61). A label permutation gave p < 0.05 in 2 of 6 patients. The group sign-flip test was not significant (p = 0.125). Stimulation pairs with an MTL contact did not cause a consistent change in neocortical N1 count (4 patients, median observed/expected 0.75, range 0.32-4.87, sign-flip p = 1.0). Thus a distance-only model often overestimates MTL responses, but one group-level MTL correction is not supported. The package contains numpy-only code, the input copy with S3 ETag provenance, results tables and an SVG figure.

hippocampus · amygdala · N1 response · CCEP · SEEG · intracranial EEG · electrical stimulation · epilepsy · distance model · brain connectivity