The patient bootstrap interval included zero added SOZ rank gain from time agreement under a fixed ridge model in 17 SEEG sleep records

We tested whether relative detector time agreement adds seizure onset zone (SOZ) rank value to a fixed baseline. The source was OpenNeuro ds005398 version 1.1.3. The main test used 17 SEEG sleep records and 474 within-shaft comparisons. The baseline used both detector counts, both marked durations, and contact position. The added feature was time Jaccard agreement. Each model excluded the test patient from fitting and scaling. Mean patient AUC rose from 0.7731 to 0.7794. The mean gain was 0.0063. Its approximate 95 percent refitted patient bootstrap interval was -0.0406 to 0.0405. This interval assumes independent, exchangeable patients and a stable fixed model procedure. All six fixed check intervals included zero. The evidence did not establish an added rank benefit under the tested procedure. It did not establish absence of benefit. This is a new analysis in a reused cohort. It does not test true event identity or clinical benefit. The package contains exact source event tables, metadata, EDF headers, code, all bootstrap draws, results, and repeat commands. All 94 source checks passed. Separate calculations matched the features, models, and 84 bootstrap draws. All ten numerical result files matched a complete repeat from fresh public bytes.

high-frequency oscillations · stereoelectroencephalography · seizure onset zone · detector agreement · patient bootstrap · ridge regression