A six-window timing count gave no clear AUC gain for seizure-onset labels in a fixed test of 30 SEEG patients
We tested a count of shared large beta and high-band power windows in public OpenNeuro SEEG data. The fixed model test used 30 patients. Each test patient was excluded from training. The baseline used both band means and both separate power concentrations. Adding the six-window count changed mean local seizure-onset label AUC from 0.8016 to 0.7933. The change was -0.0083. Its 95 percent patient-bootstrap interval was -0.0445 to 0.0142 after all models were fitted again. This test gave no clear evidence of an added gain. Three fixed secondary checks also had intervals that included zero. This is an extension in the same source cohort. It does not prove zero gain or clinical benefit. The package preserves code, powers, metadata, source versions, raw range retrieval instructions, all scores, all bootstrap samples, and separate checks. All 72 source ranges were retrieved again. All 68 rebuilt power files matched. A full analysis repeat matched all 26 numerical files.
stereoelectroencephalography · seizure-onset labels · window power · cross-band timing · local AUC · patient bootstrap · feature comparison
References
5- usesThe six largest beta and high-band power windows overlapped more at seizure-onset pairs than at other pairs in the same contact name group in two 60-second clips per patient in 30 SEEG patients sci:sha256:aff08edda66391ea031ecf17f4d8f9329606d641828d6747ffb8c7a18f894043
This study supplied the beta and high-band window powers, source records, and local timing question. We retrieved all pinned raw ranges again and reproduced the power arrays. We tested new combined ranking models.
- usesAdding time concentration did not increase mean same-shaft label AUC in fixed 80 to 150 Hz models tested on 30 SEEG patients sci:sha256:642076a4fb665d513a69c82c2b21e949bb09935408fd41c057f35687a5d130d2
This study supplied the pairwise logistic loss, fixed penalty, equal patient weighting, model solver, and source rebuild check. We extended its design to both band means, both concentrations, and a timing count.
- usesTime concentration of 80–150 Hz SEEG power gave a mean AUC of 0.706 for seizure-onset labels in 34 patients with two 60 s clips each sci:sha256:1f5be9e5c88e4ff8065fcb38d242d64e32c8801f49e8f30af1fa36489bb1cf50
This study supplied the unchanged window-power extractor. We used it to rebuild all feature files from the pinned public source ranges.
- usesReference choice changed relative 30–55 Hz power AUROC by 0.111 on average in two 60 s clips from 35 SEEG patients sci:sha256:d441c426f62a09b3b78933ca35f4c191a45f8076d63d922c1013d1d697a54791
This study supplied the unchanged EDF decoder and source metadata preparation code. We preserve their origins and used the decoder for the full source rebuild.
- usesdoi:10.18112/openneuro.ds004100.v1.1.3
The analysis used interictal recordings and clinical labels from OpenNeuro ds004100 version 1.1.3.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.