Alternating-polarity averaging separates pulse-locked SEEG responses from the 50 Hz stimulation artifact in a 2-stimulation pilot of OpenNeuro ds005169; distance-decay comparison not yet tested

This pilot asks whether pulse-locked neural responses during alternating-polarity stereo-electroencephalography (SEEG) stimulation spread with distance and decay more slowly than the stimulation artifact. A streaming pipeline separated polarity-invariant averages from polarity-flipping artifact averages and tested the neural component against a pre-stimulation surrogate null. The source was OpenNeuro ds005169 snapshot 1.0.0, which contains 4096 Hz bipolar SEEG recorded during stimulation associated with visual hallucinations. Slow downloads allowed 5 of 112 stimulation sessions to finish; polarity quality control passed for 2 stimulations from 2 subjects, yielding 69 channel observations. Median residual-to-artifact ratio during the pulse was 0.033. Eight of 69 channels exceeded the null at FDR q<0.05, including some 55–84 mm from the stimulation site. Artifact log-log distance slopes were -1.6 and -3.7 for the two subjects. Too few channels responded to estimate neural decay, and a slow (>=3 ms) artifact residue is not excluded. This pilot supports no inference about neural spread or relative distance decay; the main hypothesis remains open. Polarity detection failed for 3 sub-02 sessions. The recorded neural excess slopes are invalid, and the planned full-run analysis was not completed.

stereo-electroencephalography · SEEG · alternating-polarity stimulation · stimulation artifact · neural response · signal processing · pilot study · OpenNeuro