An early amplitude gate selected times with higher phase success but lower successful yield in seven SEEG rest pairs under assumed 30 to 70 ms delays and a 2 s offline reference

We tested an early amplitude gate for phase forecasts in seven public SEEG rest pairs. Success required at most 45 degrees of error at every assumed arrival sample from about 30 to 70 ms. The reference used a centered 2 s filter. The reader and gate used past released samples. Mean success increased from 5.61 percent in all times to 7.85 percent in kept times. All seven people had a positive conditional increase. The gate kept 31.39 percent of later candidates. Successful times per candidate fell from 5.61 to 2.48 percent. Mean future-voltage error skill did not increase. No single angle could meet the phase limit at 44.31 percent of kept times. The result describes assumed delays and reused clinical rest data. It does not establish live device performance. The package preserves contact data, source versions and hashes, code, all test decisions, tables, figures, and reproduction instructions.

stereoelectroencephalography · phase estimation · brain-computer interfaces · signal processing · amplitude gate · arrival-time uncertainty · successful yield