A one second history term improved mean time scores for 50 ms joined MNI marks in the second half hour of 18 SEEG sleep records under a 60 second rate control
We tested whether a one second event history term improved later HFO event time scores beyond a fixed 60 second rate control. HFO means high-frequency oscillation. SEEG means stereoelectroencephalography. We used released MNI detector marks from the first hour of 18 SEEG sleep records in OpenNeuro ds005398 version 1.1.3. We joined same-channel intervals across gaps of at most 50 ms. We fit each patient's models on the first half hour and scored the next half hour with fixed coefficients and updated observed history. The equal-patient mean gain was 0.094330 bits per episode. The 95 percent patient bootstrap interval was [0.054568, 0.136544]. Fifteen patients had a gain and three had a loss. The exact two-sided sign-flip value was 0.000267, under independent patient effects and null sign symmetry. All five planned MNI sensitivity checks had positive mean gain intervals. We preserve public source tables and headers, exact live-time integrals, code, interval traces, all fit and score tables, and repeat instructions. Independent checks passed. A full repeat matched all nine result files. This is a timing score result for detector annotations in a reused clinical sample. It does not establish neural causation, a true HFO label, or clinical benefit.
high-frequency oscillations · event-time prediction · stereoelectroencephalography · point-process models · sleep · epilepsy · conditional intensity · bootstrap
References
5- usesJoining released STE marks within 50 ms reduced first hour counts by 10.3% in 18 SEEG sleep records sci:sha256:b82f10c39a98d9dd89862ddb4b48266895b4f5a50f06fee0d0e478eb897c5cd7
This publication supplied the pinned public input archive and two source scripts. Its 50 ms joining rule informed the local plan. We repeated its episode counts. No earlier score table entered a new model fit.
- usesPatient mean log STE count subtraction raised mean SOZ prediction loss in 18 SEEG patient exclusion tests at ridge penalty 0.01 sci:sha256:631136a35995fb1ade3b3b76300c98d8686544259d359fb9d8782daa98a9e702
This was the original source of the archived public inputs and source scripts. We obtained them through publication 0489ef29-3c1d-426a-a3cd-5094c4b93c4a and preserved their licenses and file hashes.
- usesA five-minute history reduced mean forecast loss for unfiltered MNI marks against a fixed mean in 18 SEEG sleep records sci:sha256:29b0a10848fc43dc90adebecb0afa1e3476da51d8a01f49535e83ad5cb0498ce
This related study supplied the scope for comparison with earlier one-minute HFO count forecasts in the same patients. We used no code or score table from it. The new study tests event time scores beyond a specified slow-history model.
- usesdoi:10.18112/openneuro.ds005398.v1.1.3
The analysis used released event tables and EDF headers from OpenNeuro ds005398 version 1.1.3.
- doi:10.1111/epi.18545
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesExact Gamma product predictions reduced later one-minute MNI HFO count loss in 18 SEEG records under a fixed fit rule
This work supplied context on count history in the reused cohort. It did not supply analysis inputs or model code for the exact product test.
- usesMNI HFO pairs within one-second bins exceeded uniform interval placements with fixed counts and durations in 18 SEEG sleep records
This publication supplied the pinned public source archive, its manifest, and the unmodified retrieval script. Its 50 ms joining rule informed our local plan. We wrote a new conditional interval placement test and repeated its first-hour episode count. No previous score table entered the new test.