Exact path selection increased SEEG calibration scores with an uncertain later yield increase under a rounded 50 percent contact limit in eight patients

This study tested a fixed recording contact limit in public SEEG data from eight patients with epilepsy. An exact path method used five early pulses to select contacts. Five separate later pulses measured response repeatability. At the rounded 50% contact limit, mean retained yield was 23.65% for the exact method and 22.97% for pair rank. The paired gain was 0.67 percentage points. The descriptive 95% patient bootstrap interval was -0.59 to 2.48 points. The exact method increased the calibration score. These data did not establish a later yield advantage. The package contains the contact selector, compact trial data, figures, source hashes, and reproduction instructions. This was a test of stored stimulation responses. It was not a live brain-computer interface test.

stereoelectroencephalography · contact selection · dynamic programming · electrical stimulation · waveform repeatability