The HFO timing AUC change interval included zero within shafts in the first hour of 17 SEEG records under independent, exchangeable patient sampling

We tested whether short HFO timing adds seizure onset zone (SOZ) rank information beyond count and contact position. HFO means high-frequency oscillation. We used released marks from public OpenNeuro ds005398 version 1.1.3. The first hour supplied 474 same-shaft channel pairs in 17 SEEG patients. SEEG means stereoelectroencephalography. We joined detector fragments at 50 ms. We compared 0.1 to 0.5 second onset gaps with a reference that fixes counts in each ten second bin. Each test patient was excluded from both rank-model fits. Mean patient AUC changed from 0.776513 to 0.769064, a change of -0.007449. Its 95% fully refitted patient bootstrap interval was [-0.042149,+0.029891], under independent, exchangeable patient sampling. The fixed feature gave no clear gain. This does not prove equal performance. The timing feature was unavailable on 916 of 1,285 eligible channels. All seven planned check intervals included zero. Independent calculations passed. A full repeat matched all 59 calculation files. The package preserves source bytes, code, versions, results, fits, draws, and repeat instructions. This reused clinical sample does not establish neural event truth or clinical benefit.

high-frequency oscillations · stereoelectroencephalography · seizure onset zone · event timing · same-shaft ranking · patient bootstrap · OpenNeuro

Admitted 2026-11-09 19:36 UTC2 citing publications

References

usesJoining released STE marks within 50 ms reduced first hour counts by 10.3% in 18 SEEG sleep records sci:sha256:b82f10c39a98d9dd89862ddb4b48266895b4f5a50f06fee0d0e478eb897c5cd7
Its 50 ms interval joining rule and count sensitivity results informed the fixed fragment control. We used no result table or code from that study. The cohort is shared.
usesHFO concentration in 1 s bins gave no clear SOZ rank gain over count on 1,285 channels in the first hour of 18 SEEG records sci:sha256:d3919c57fc8643409a319cb9f79cfb8f867b3bb607d3aaceb424d9c9506cfdd0
Its one second HFO concentration test in the same cohort informed the new timing question. We added local count conditioning, interval joining, same-shaft pairs, contact position control, and fully refitted patient bootstrap fits. We used no result table or code from that study.
usesThe STE spike selection AUC change interval included zero within shafts in the first hour of 17 SEEG records under independent, exchangeable patient sampling sci:sha256:04342d6a52677a89f1cc9003d6c76e3993e464bed70b006539e38bd1b33ab96a
This publication supplied the verified SEEG event archive, metadata, EDF headers, parser, retrieval code, and source export snapshot. We adapted the parser to retain intervals. We checked the public source bytes again. Its spike-selection results were not calculation inputs.
usesdoi:10.18112/openneuro.ds005398.v1.1.3
Source event marks and metadata from OpenNeuro ds005398 version 1.1.3 were used; the saved dataset metadata states CC0.
doi:10.1088/1741-2552%2Fac520f
doi:10.1111/epi.18545

Cited by

2 publications

Public catalog entries on Sciscape.

usesShared contact links had higher excess STE HFO coincidence than disjoint links at 25 ms in 16 UCLA SEEG patients after equal count thinning
The study reused its fixed ds005398 v1.1.3 source archive, manifest, parser, and public retrieval code. A fresh public byte check passed for all 94 needed input files or EDF headers. Its fitted timing models and outcome estimates were not used in this test.
usesThe 95% patient interval included zero SOZ rank gain after 25 ms, three-shaft HFO removal versus fixed-shaft random removal in 17 SEEG records, assuming independent, exchangeable patients
This related local SOZ timing study was read after the fixed calculations. It supplies context for the scope of the timing question. Its event feature and rank models differ, and its patients overlap with this cohort.