Early pulse calibration selected eight sites with higher later SEEG identification accuracy than random sets within site blocks in five patients

We tested whether early pulses can select stimulation sites with distinct later SEEG responses. We used 1,923 pulses at 185 sites in OpenNeuro ds004457, version 1.0.2. The first half of pulses at each site supplied calibration data. The second half supplied separate test data. A fixed rule selected eight sites for each of five patients. Mean site identification accuracy was 82.1 percent. Uniform random sets reached 41.8 percent. The gain was positive in all five patients. Its descriptive patient bootstrap interval was 30.9 to 50.6 percentage points. The exact two-sided patient sign-flip p value was 0.0625. Selecting the strongest responses reached 79.6 percent. The evidence did not show a clear gain over that simpler method. Signals before the pulse reached 11.1 percent for the selected sets, against 12.5 percent chance. The package preserves signals, code, all sets, predictions, three figures, source hashes, and raw retrieval instructions. Complete raw extraction and separate scalar checks passed. The test halves shared stimulation blocks and sessions. Calibration was earlier within each site, but not earlier than all test pulses across the recording. The result does not establish new-session performance, distinct percepts, or useful clinical control.

stereoelectroencephalography · electrical stimulation · site identification · calibration · brain-computer interfaces