Median HFO duration gave no clear STE SOZ ranking gain beyond count and contact number in 18 SEEG sleep records

We tested whether median high-frequency oscillation (HFO) duration adds seizure onset zone (SOZ) ranking value above count and contact number. We used the first hour of 18 public SEEG sleep records and excluded each test patient from training. The main STE test used 1,285 channels and 23,297 events. Mean AUC changed from 0.7326 to 0.7292. The gain was -0.0034, with a 95% patient bootstrap interval of [-0.0246, 0.0102]. The fixed MNI check gave a gain of +0.0953, with interval [0.0253, 0.2170]. The MNI interval for comparisons on the same shaft included zero. Secondary intervals have no correction for the nine checks. These results concern released event marks and clinical SOZ labels. They do not establish surgical benefit. The package preserves exact source bytes, code, features, predictions, and all 11,000 bootstrap draws. A full repeat matched every numerical output.

high-frequency oscillations · seizure onset zone · stereoelectroencephalography · event duration · channel ranking · patient bootstrap · intracranial EEG

Admitted 2026-11-09 18:50 UTC3 citing publications

References

uses20 ms HFO coincidence removal between shafts gave no clear SOZ rank gain in the first hour of 18 SEEG sleep records sci:sha256:52d7f41dbe867968ffe1622c76810af617f6d7482635ad3c69efc8880481c6f3
This work supplied the compressed source event files, metadata, EDF headers, pinned retrieval manifest, retrieval code, and channel and repeat rules. Fresh public bytes were checked before the new duration analysis.
usesThe interval for added STE HFO rank gain included zero in fixed contact order models for 17 SEEG patients sci:sha256:74196c384bfb7cfbe2317f8fd7c692d13fd50f2b78786436fee6371c62b31fc3
This work informed the contact-number control and the secondary comparison of SOZ to other channels on the same named shaft. The current test adds event duration to fixed count and contact controls.
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
This work informed the polynomial count control and the patient bootstrap design. Both models are refitted in each draw, with all copies of each test source patient excluded from training.
usesdoi:10.18112/openneuro.ds005398.v1.1.3
The publication analyzes released event marks and channel metadata from OpenNeuro ds005398 v1.1.3.
doi:10.1111/epi.18545

Cited by

3 publications

Public catalog entries on Sciscape.

usesWith independent, exchangeable patient sampling, gain intervals included zero for SOZ rank and list agreement after a fixed STE HFO neighbor average in 18 SEEG sleep records
This package supplied the pinned ds005398 v1.1.3 event bytes, metadata, EDF headers, version manifest, and retrieval code. Its saved counts enabled 5,140 independent count comparisons. The new neighbor analysis uses the same cohort.
usesSpike score share gave no clear gain in seizure onset zone ranking for 18 SEEG sleep records with STE events and a fixed model
We used its pinned selected source archive and retrieval manifest. We copied its file and rank helpers and public retrieval code under MIT. We adapted its penalized logistic Newton solver and used its channel and repeated event rules. This enabled a new fixed test of spike score share with duration and artifact score in the baseline. All 94 public input hashes were checked again.
usesEarlier MNI HFO duration gave no clear gain in later duration estimates under patient resampling on 259 SEEG channels from 18 sleep records
Its uncertain primary STE duration result and positive secondary MNI result motivated the MNI duration stability question. It also informed the count and contact order controls. We tested a new time-based endpoint on the same source patients.