Five consecutive SEEG decisions did not show higher mean response detection under a shared calibration limit in eight sessions

We tested whether five consecutive decisions above a threshold improve a fixed SEEG decoder. We used public OpenNeuro ds004473 version 1.0.2 and one session from each of eight patients. Each patient had epilepsy. Each rule had its own threshold. Earlier continuous data set each threshold under the same limit of 10 background commands per minute. Separate later data gave mean response detection of 22.8% for one decision and 18.5% for five decisions. The paired change was -4.3 percentage points. Its descriptive 95% patient bootstrap interval was -14.0 to +3.7 points. An exact paired test by sign reversal gave p = 0.453. Mean background command rates were 7.2 and 7.5 per minute. At a shared zero threshold, five decisions reduced both background commands and detection. This offline result did not establish a detection gain or equal performance. Background commands are a proxy for unintended commands. The study did not test online control or brain stimulation. The package contains source code, fixed features, source versions, checks, tables, and four figures.

stereoelectroencephalography · brain-computer interface · response detection · persistence rule · calibration · offline replay