Pre-movement high-gamma activity did not exceed chance in predicting confidence in eight SEEG patients

This group analysis extended a one-subject pilot of the MetaRDK random-dot task to eight patients with stereo-electroencephalography (SEEG). It tested whether mean high-gamma activity (HGA) in the 0.5 s before choice movement predicted confidence after accounting for accuracy, log response time, and stimulus coherence. Across 1,618 analyzed trials and 714 bipolar channels, 33 channels had p < 0.05; the within-subject residual-permutation null had a mean of 35.1 and a 95th percentile of 60 (p = 0.53). The primary hypothesis was not supported. Without the pilot patient, 30 of 664 channels were significant (p = 0.51). The reported stimulus-window count was 41 of 714 (p = 0.28); this was not a tested hypothesis. The insula analysis found no group effect (seven patients, mean t = 0.18, sign-flip p = 0.60, Holm p = 1.0); the PMdm result included only two patients and was not interpretable. The analysis covered 8 of 21 patients, selected in ascending derivative-file size before a run deadline. It used window means only, and the channel-count test had low power for sparse or weak effects. The analyzed files came from the public OpenNeuro ds006253 derivatives URL; the exact dataset release is uncertain.

stereo-electroencephalography · high-gamma activity · confidence · decision making · random-dot task · permutation test · MetaRDK