Thresholds from other patients gave a 10.2% mean background call rate at a 5% target after five fit pulses at 86 selected SEEG sites in eight patients

We tested whether background samples from other patients can set a low response threshold after five local fit pulses. We used public SEEG data from OpenNeuro ds004696 version 1.0.1. The test used 354 later pulses at 86 fixed sites in eight patients. At a 5% target, thresholds from seven other patients gave a 10.23% mean background call rate. A five-sample Student rule gave 8.29%. Response call rates were 81.10% and 77.75%. The means give each patient equal weight. The donor minus Student background difference stayed positive in all eight patient deletion checks. Removing almost duplicate source events gave the same direction. The methods used the same scores, so threshold choice did not improve response ranking. The donor rule did not establish the intended low call rate in this cohort. Its usual error bound needs exchangeable calibration and future scores. That condition was not established. Before-pulse controls do not measure every source of false calls after stimulation. The package contains figures, selected voltage, event maps, fitted models, all scores, source hashes, retrieval code, and analysis instructions. All primary calls passed an independent calculation. A repeated analysis reproduced all 12 numerical artifacts byte for byte.

SEEG · stereoelectroencephalography · threshold calibration · brain-computer interface · evoked response · background activity · calibration transfer · patient-level evaluation