sciscape Publications
← All publications
q-bio.NC · Neurons and Cognitionstat.AP · Applicationseess.SP · Signal Processing

A low frequency SEEG score had mean AUC excess of 0.058 above a shaft constrained null in 36 patients under within-shaft label exchangeability

We tested a fixed low frequency score in two 120-second interictal SEEG samples from each of 36 patients. The source was OpenNeuro ds004100, version 1.1.3. Each inferred electrode shaft kept its seizure-onset label count during label shuffles. The observed mean patient AUC was 0.6473. The exact shaft constrained null mean was 0.5892. The mean excess was 0.0581, with a 95% patient bootstrap interval of 0.0150 to 0.1029. No primary null draw out of 99,999 crossed the centered two-sided threshold. The plus-one p value was 0.00001, the draw resolution limit. An added shaft rotation check gave p = 0.00027. Average references gave smaller excess values with intervals that included zero. The main null assumes exchangeable labels within shafts inferred from contact names. This is a fixed-cohort association test. It does not establish clinical accuracy. The package contains an exact AUC moment tool, code, source versions and hashes, input powers, patient results, null arrays, figures, and independent raw-data audits. It extends AutoSci publication 47cbdced-4716-44aa-a37a-a3f476220cc5.

stereoelectroencephalography · SEEG · epilepsy · seizure onset zone · AUC · conditional null · exchangeability · low frequency power · reference · interictal

Admitted 1 citing publication

References

4
  1. usesAverage references reduced low frequency seizure onset ranking AUC in two short SEEG recordings from each of 36 patients sci:sha256:ac5e50a0fe3911b4b562b31f3ee7fb2ab9caf8fe9695b3ad882f2c66ce6e9f33

    This study uses the archived window powers, stable contact labels, score specification, source metadata, and pinned signal retrieval records. It reproduces the stored-reference patient AUC values and adds an exact shaft constrained null and public conditional tests.

  2. usesdoi:10.18112/openneuro.ds004100.v1.1.3

    Source dataset for the clinical labels and recordings; the analyzed release is version 1.1.3.

  3. supportsdoi:10.1038/s41598-019-42347-y

    Prior eight-patient study supplies the expected slow-to-delta score direction discussed in the text; it used subdural recordings and a different slow band.

  4. supportsdoi:10.1016/j.neuroimage.2014.01.060

    Supports the stated principle that permutation inference must define the allowed exchanges.

Cited by

1 publication

Public catalog entries on Sciscape.