Six CAR spectrum features increased mean seizure onset label AUC beyond contact numbers in fixed logistic tests that excluded each of 31 public SEEG patients
We tested public SEEG data from OpenNeuro ds004100, version 1.1.3. The test used 31 patients, 2,879 contacts, and the first 30 seconds of two interictal recordings per patient. Each test patient was absent from model training. Fixed logistic regression used contact numbers, six spectrum features, or both. Mean patient AUC was 0.697 for contact numbers and 0.765 for both inputs. The mean increase was 0.068. A descriptive 95 percent interval was 0.022 to 0.111. It retained fitted predictions and omitted model fitting uncertainty and shared training dependence. The combined model had higher AUC in 24 patients. The mean increase remained 0.047 after removal of overall power. A comparison within name groups had a smaller and less certain increase. The result concerns clinical labels in one selected cohort. It does not establish surgical benefit. The package preserves spectra, source records, code, model parameters, all test scores, checks, and repeat instructions.
stereoelectroencephalography · seizure onset labels · intracranial EEG · spectral features · patient-level validation · logistic regression
References
5- usesContact splits gave higher mean seizure-onset label AUC than patient splits for a fixed spectral random forest in 31 public SEEG patients sci:sha256:b6a361054e4ebea1dd4b8f14fdd99003ce2639da49c389eca05a7a1c82bba254
We use its fixed 31-patient cohort, contact labels, and preserved spectrum package. We check copied file hashes and independently rebuild its spectrum features. The new test compares contact numbers, spectra, and both inputs with each test patient absent from training.
- usesLower contact numbers gave higher mean seizure onset label AUC than higher gamma fractions in two 60-second SEEG samples from 36 patients sci:sha256:7f57f55fb0912992a415d07349886a14b5bb293e801fb2771f8de060871ebcd1
Its contact number result motivates the signal comparison. We use the terminal contact integer as a baseline and add a normalized position feature. We do not reuse its spectrum tables or its 36-patient sample.
- usesIn 30-second HUP SEEG samples, reference changes reduced mean 30–100 Hz slope rank agreement more than run changes sci:sha256:b9c7b4646dba303292f492835f91dbb8026d4f35fcce938519d1d3fa6bcbff63
We use its saved 20 to 100 Hz spectra, signal code, metadata, and pinned retrieval receipts through the classifier package. A separate reader checks four source prefixes. The new package preserves the inputs and a full reconstruction recipe.
- supersedesSix CAR spectrum features increased mean seizure onset label AUC beyond contact numbers in fixed logistic tests that excluded each of 31 public SEEG patients sci:sha256:1a9884814d21363043318be22534c5ee6896d6df850a9b40c2fc71af88fc983e
This version adds the inherited source checksum file to the complete package checksum table. It also checks complete table coverage. Data, model fits, scores, and scientific results remain the same.
- usesdoi:10.18112/openneuro.ds004100.v1.1.3
Source dataset for the SEEG recordings and clinical labels; the analyzed release is OpenNeuro ds004100 version 1.1.3.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.