A complete power bound prevented releases with error above one percent of early mean power after 10 ms gaps filled with zero in 35 SEEG records from 21 patients under a missing voltage cap

We tested a power error gate after missing SEEG voltage. The study used 35 public records from 21 patients and four fixed channels per record. Each simulated gap was filled with zero. The main test used nominal 10 ms gaps, causal 70 to 110 Hz band voltage, and a 100 ms power mean. The error limit was one percent of an earlier mean power. A proof gave a bound for the complete feature under an absolute cap on missing voltage and equal state before the gap. All main missing samples met the selected cap. The complete gate had no above-limit releases in the 500 ms output interval. A gate for the current squared signal had a bad release rate of 0.465 percent of all output opportunities. The descriptive patient bootstrap interval for the paired reduction was 0.358 to 0.578 percentage points. The complete gate accepted 34.926 percent of outputs; the current gate accepted 49.387 percent. Each patient had equal weight. The package contains the proof, streaming reference code, all selected voltage, source hashes, public retrieval code, counts, checks, and four figures. All 35 selected records matched a fresh source rebuild. A second analysis produced 16 byte-identical result files. The voltage cap is an external assumption. This test did not measure decoding, live stimulation, or clinical safety.

causal signal processing · power error bound · missing data · stereoelectroencephalography · SEEG · brain-computer interfaces · streaming algorithms