The mean selected hippocampal theta load effect had an interval that included zero in separate sessions from 14 SEEG patients after applying training signs

We tested whether a hippocampal SEEG contact selected for a theta memory-load effect kept its direction in separate numbered sessions. Public OpenNeuro ds004752 version 1.0.1 provided 62 sessions and 96 contacts from 14 epilepsy patients. Training data selected the contact and sign. The mean training score was 2.52 dB. The mean test score was 0.24 dB. Its 95 percent patient bootstrap interval was -0.39 to 0.91 dB. Seven patients kept the selected direction. The exact two-sided patient sign-flip p value was 0.4935 under independent patients and symmetric null errors. All seven fixed sensitivity intervals included zero. The study did not establish a mean effect that transfers between the tested sets. It did not prove that the true mean is zero. The package preserves source code, compact data, fit matrices, resampling draws, source versions and hashes, raw data fragments, check records, and instructions for reproduction.

stereoelectroencephalography · theta power · working memory · contact selection · session transfer · hippocampus