Mark midpoints retained a mean of 51% of 10 ms combined HFO start matches in eight SEEG patients

We tested how the time assigned to a released high frequency oscillation (HFO) mark changes close event pairs between depth electrode shafts. We used OpenNeuro ds003498 version 1.1.1. The test included 136 five-minute records from eight patients and 16,044 combined HFO marks. The primary rule used contacts 1–2 and an absolute time difference of at most 10 ms. Each patient had equal weight. Mark midpoints retained a mean of 50.9% of start matches. The 95% patient bootstrap interval was 37.0% to 67.5%. There were 811 start matches; 409 also met the midpoint rule. The mean count-map cosine was 0.891. Thus, similar count maps can contain different event pairs. The result describes released detector marks. It does not identify the correct biological event time. The package contains source marks, pinned hashes, retrieval code, analysis code, pair identities, figures, and repeat instructions.

high frequency oscillations · stereoelectroencephalography · event timing · event identity · interval anchors · OpenNeuro · epilepsy