A fixed shape gate did not establish higher 10 ms peak agreement in 140 selected SEEG pairs

We tested a fixed waveform-shape gate for single-pulse feedback after electrical brain stimulation in eight clinical SEEG sessions. Early calibration selected 140 recording pairs at 60 stimulation sites from OpenNeuro ds004696 v1.0.1. The gate accepted a later waveform when its correlation with the five-pulse calibration mean was at least 0.8. In a 20–500 ms window, a voltage peak agreed when it fell within 10 ms of the calibration peak. With equal patient weight, yield was 70.5%, and agreement among accepted outputs was 73.0%, compared with 69.5% without the gate. The paired change was 3.6 percentage points, with a descriptive 95% patient bootstrap interval from −1.4 to 8.2 points; the main test therefore did not establish higher agreement. The gate withheld agreeing readouts equal to 16.5% of all attempts. Mean acceptance of before-pulse controls was 0.45%. Some accepted waveforms had peak shifts above 450 ms despite high shape correlation. A fresh download reproduced all eight compact voltage files exactly. This exploratory retrospective test measures voltage-peak repeatability in selected clinical data. It does not test live brain-computer-interface performance.

SEEG · brain stimulation · single-pulse feedback · waveform correlation · peak timing · output yield · selective readout · reproducibility