Noise correction of centered 20 to 500 ms voltage changed the selected SEEG pair at 15 of 84 sites in eight sessions after five early pulses

We tested selection of one SEEG recording pair for response feedback after brain stimulation. Public OpenNeuro ds004696 version 1.0.1 supplied eight epilepsy sessions, 84 sites, and 884 pair instances. Five early good pulses set each choice. Five separate final good pulses measured repeated voltage energy at 20 to 500 ms. Removal of estimated noise from mean-waveform energy changed 15 choices. Eleven changed choices had a higher final score and four had a lower score. The equal-patient mean difference was +0.0582 units of early maximum mean-waveform energy. The 95 percent patient bootstrap interval was 0.0055 to 0.1248. The exact two-sided patient sign-flip test gave p = 0.125. Selection with two or three early pulses had a lower mean score. The study does not establish a population or live-interface gain. The package preserves exact inputs, a calibration selection tool, source hashes, complete result tables, three figures, and replay instructions. Fresh complete source rebuilds for patients 01 and 08 matched all archived arrays and file bytes exactly. The cohort and final pulses had been used in prior studies.

stereoelectroencephalography · electrical stimulation · noise correction · pulse selection · brain-computer interface · repeated voltage energy