A four-group reciprocal shape contrast exceeded its pre-pulse control at 20 to 500 ms in seven SEEG sessions

A brain interface can use measured stimulation responses to build a map. We tested whether a scalar can represent the reverse response shape. We used public OpenNeuro ds004696 version 1.0.1, with 170 exact reciprocal bipolar pairs in seven patients. Four separate pulse groups estimate a shape-area moment. Two other pulses per direction set positive weights. The mean post-minus-pre contrast was 0.220. A descriptive 95 percent patient bootstrap interval was 0.079 to 0.424. All seven patient contrasts were positive. The expected moment is zero for proportional mean waveforms under stable means and independent, mean-zero group errors. A shared-error simulation gives a false positive contrast. The recordings do not establish those assumptions. The result concerns measured voltage and does not prove neural directionality or brain-interface performance. This work adds a noise check to an existing cohort analysis. It adds no new patients. The package contains proof, source recipes and hashes, code, audit voltage, results, and five figures. The GPL analysis and retrieval scripts are omitted from SCI, so the exact analysis implementation is not included for public replay.

stereoelectroencephalography · electrical stimulation · reciprocal responses · waveform shape · independent-repeat estimator · OpenNeuro ds004696