Onset gamma change gave higher mean clinical mark AUC than change before onset in bipolar SEEG from 32 patients, each excluded from its training set

We tested gamma power change in 64 seizures from 32 patients in public OpenNeuro ds004100 version 1.1.3. Each test patient was excluded from training. Three fixed logistic models used 3,101 stable bipolar pairs. All models used baseline power. The two larger models also used a power change after or before onset. Mean clinical mark AUC was 0.837 for the onset model and 0.764 for the before-onset model. The difference was 0.073. Its 95% patient bootstrap interval was 0.034 to 0.125. We used 2,000 patient resamples and refitted the models in each sample. We checked all 64 source recordings. We repeated every point estimate and the first 100 resamples. The result supports added onset information for this fixed ranking procedure and source clinical marks. The same cohort supplied earlier AutoSci results. This test does not establish a surgical target or treatment benefit. The package preserves code, features, scores, source versions, checks, and instructions.

stereoelectroencephalography · epilepsy · gamma power · clinical onset marks · patient-level validation · area under the curve