Two six-contact SEEG rules gave the same later site coverage in eight limbic sessions at a 0.8 shape cutoff over 20–500 ms

This study compared two exact recording-selection rules in eight public limbic SEEG sessions from OpenNeuro ds004696 version 1.0.1. Each rule chose three bipolar pairs with six different contacts. One rule first maximized covered stimulation sites. The other first maximized accepted site-pair responses. The first five good pulses supplied selection data. The final five supplied separate test data. At a Pearson shape cutoff of 0.8 over 20–500 ms, both rules covered the same later sites in every person. Mean later site coverage was 73.63% for both rules. Seven people had the same selected pairs. The remaining person had different pairs but the same ten later covered sites. The exact random-set mean was 56.83%, and the later-data upper bound was 76.26%; these controls are descriptive. The package preserves all 1,635 valid six-contact sets, exact trial products, all planned checks, four figures, code, source hashes, and replay instructions. Independent code checked every score and selected set. Three full public source files were rebuilt and matched exactly. This is an exploratory offline within-site test. It does not establish population equivalence, neural origin, or prospective brain-computer interface performance.

stereoelectroencephalography · site coverage · recording selection · waveform correlation · brain-computer interface