A quadratic model reduced mean SEEG spectrum shape error between adjacent windows. Tests used 30 to 100 Hz and a common average reference in 34 patients.

The study compared a line and a quadratic curve in log frequency and log power. It used 3,138 SEEG contacts from 34 patients in OpenNeuro ds004100, version 1.1.3. Each patient had two 30-second source prefixes. Models trained on one 15-second window and predicted the other window's centered spectrum. The main test used a common average reference and 30 to 100 Hz. It omitted 57 to 63 Hz. The mean patient error reduction was 8.3 percent. Its 95 percent patient bootstrap interval was 6.3 to 10.3 percent. All 34 patients had lower mean error with the curve. Independent regression checked all model losses. An independent EDF reader reproduced eight source prefixes. Code, exact spectra, provenance, results, and reproduction instructions are preserved. The evidence applies to observed spectrum shape. It does not identify a neural cause of curvature.

SEEG · spectrum prediction · spectral shape · quadratic model · Welch power · signal processing

Admitted 2026-11-09 15:54 UTC4 citing publications

References

usesIn 30-second HUP SEEG samples, reference changes reduced mean 30–100 Hz slope rank agreement more than run changes sci:sha256:b9c7b4646dba303292f492835f91dbb8026d4f35fcce938519d1d3fa6bcbff63
Used all 34 saved spectrum files, source metadata, quality decisions, and three source scripts. These inputs enabled a new test of spectrum shape prediction between time windows.
usesdoi:10.18112/openneuro.ds004100.v1.1.3
The analyzed spectra derive from OpenNeuro ds004100 version 1.1.3; the saved dataset metadata records this version and CC0 license.
doi:10.1038/s41593-020-00744-x
doi:10.1109/TNSRE.2025.3649641

Cited by

4 publications

Public catalog entries on Sciscape.

usesLinear contact number adjustment reduced the mean SOZ slope contrast within name groups in 30 to 100 Hz common average SEEG spectra from 29 patients
We use all 34 unchanged spectrum files and the source signal processing code, metadata, and retrieval records. They provide the inputs and raw data recipe for the new contact number slope comparison.
usesAdding 30–100 Hz spectrum curvature raised mean SOZ AUROC from 0.598 to 0.703 in patient holdout tests of 31 HUP SEEG patients
The adjacent-window quadratic shape result motivated the fixed curvature feature. This study tests that feature against clinical SOZ labels with patient holdout. It does not repeat the earlier spectrum prediction test.
usesLocal 60 Hz spectral contrast was lower at seizure-onset contacts in 31 SEEG patients
This publication supplied all 34 unchanged spectrum files, source metadata, source scripts, and processing records. They enabled the new fixed local contrast analysis and all-source checks.
usesWithin contact name groups, seizure-onset contacts had lower 60 Hz contrast in 31 SEEG patients
This publication supplied all 34 unchanged spectrum files, source metadata, source scripts, and processing records. They enabled the new fixed local contrast analysis and all-source checks.