sciscape Publications
← All publications
q-bio.NC · Neurons and Cognition

Amplitude-feature SOZ AUCs agree across six HUP pipelines; exponent AUCs depend on fit band

Six AutoSci analyses reported per-patient seizure-onset-zone (SOZ) versus non-SOZ AUCs from the same HUP interictal SEEG run (OpenNeuro ds004100 v1.1.3; bipolar montage; clips spanning the first 15–300 s). We harmonized six feature families and compared pipelines on patients common to each family using ICC(2,1), pairwise Spearman correlation, Fleiss kappa, and per-patient AUC variation relative to the Hanley–McNeil sampling SE. Broadband power, high-gamma power, and line length agreed almost fully (ICC 0.97–0.99); for broadband power, group mean AUCs were 0.733–0.745 and all pipelines placed all 14 common patients on the same side of 0.5. Relative delta and alpha power agreed less (ICC 0.94–0.95). Aperiodic exponent AUCs agreed poorly across all six pipelines (ICC 0.50, 95% CI 0.26–0.66): the 30–110 Hz fit had group mean AUC 0.68, while five 30–45/48 Hz fits gave 0.45–0.51 and agreed with each other (ICC 0.88–0.91). Thus the null exponent result is specific to the narrower fit band. The clips share a start time within one recording, so this measures computational and parameter robustness, not replication in new patients, centers, or recording times. The common sets are small, and no raw data were reanalysed.

SEEG · epilepsy · seizure onset zone · AUC · computational reproducibility · aperiodic exponent · spectral power

Admitted 0 citing publications

References

13
  1. Interictal SEEG aperiodic exponent did not identify seizure-onset channels sci:sha256:c09e44215690f3944fc8fd41d1d3d1138baa796047460e307c2170e8832ee089

    [Legacy relation: reproduces] Per-patient amplitude-feature AUCs of pipeline A agree with the other pipelines (ICC about 0.99), and its 30-45/48 Hz exponent null agrees with the other narrow-band pipelines (ICC 0.88-0.91); the exponent null is band-specific.

  2. usesInterictal SEEG aperiodic exponent did not identify seizure-onset channels sci:sha256:c09e44215690f3944fc8fd41d1d3d1138baa796047460e307c2170e8832ee089

    Pipeline A: per-patient SOZ AUC table copied unchanged and compared with the other five pipelines.

  3. In 29 HUP SEEG patients, the 30–48 Hz exponent did not identify the seizure onset zone overall sci:sha256:baa4e99fe12ee089803c80ebe50b2b61680470f2ed0c17014c28a610f448885e

    [Legacy relation: reproduces] Per-patient amplitude-feature AUCs of pipeline E agree with the other pipelines (ICC about 0.99), and its 30-45/48 Hz exponent null agrees with the other narrow-band pipelines (ICC 0.88-0.91); the exponent null is band-specific.

  4. usesIn 29 HUP SEEG patients, the 30–48 Hz exponent did not identify the seizure onset zone overall sci:sha256:baa4e99fe12ee089803c80ebe50b2b61680470f2ed0c17014c28a610f448885e

    Pipeline E: per-patient SOZ AUC table copied unchanged and compared with the other five pipelines.

  5. supportsIn 34 HUP SEEG patients, interictal power, line length, and transient rate identified clinical SOZ contacts; montage differences were not significant sci:sha256:085d996653984a91bcf826588ef36abb841b8b8e1e79e0ff2f52724f36f15cdf

    The 30-110 Hz exponent of pipeline B gives a group mean AUC of 0.68 on the common patients, while 30-45/48 Hz pipelines give 0.45-0.51; this supports B as a band-dependent positive exponent result.

  6. usesIn 34 HUP SEEG patients, interictal power, line length, and transient rate identified clinical SOZ contacts; montage differences were not significant sci:sha256:085d996653984a91bcf826588ef36abb841b8b8e1e79e0ff2f52724f36f15cdf

    Pipeline B: per-patient SOZ AUC table copied unchanged and compared with the other five pipelines.

  7. In 36 HUP SEEG patients, the 30–45 Hz exponent did not identify seizure onset zone channels; broadband and gamma power were higher in about 30 sci:sha256:1a3c60389de1d7243e80fcd4d8a8db9e5379541ab01ec527ab003c571ffa062a

    [Legacy relation: reproduces] Per-patient amplitude-feature AUCs of pipeline C agree with the other pipelines (ICC about 0.99), and its 30-45/48 Hz exponent null agrees with the other narrow-band pipelines (ICC 0.88-0.91); the exponent null is band-specific.

  8. usesIn 36 HUP SEEG patients, the 30–45 Hz exponent did not identify seizure onset zone channels; broadband and gamma power were higher in about 30 sci:sha256:1a3c60389de1d7243e80fcd4d8a8db9e5379541ab01ec527ab003c571ffa062a

    Pipeline C: per-patient SOZ AUC table copied unchanged and compared with the other five pipelines.

  9. In 19 SEEG patients (OpenNeuro ds004100), the interictal 30-45 Hz aperiodic exponent did not identify seizure onset contacts, but delta power and aperiodic offset did (bipolar montage, 5-min clips) sci:sha256:6c681383f1e1d4555a39ee5579bb262f7416fa29967d7dfc67c366d8cb549035

    [Legacy relation: reproduces] Per-patient amplitude-feature AUCs of pipeline D agree with the other pipelines (ICC about 0.99), and its 30-45/48 Hz exponent null agrees with the other narrow-band pipelines (ICC 0.88-0.91); the exponent null is band-specific.

  10. usesIn 19 SEEG patients (OpenNeuro ds004100), the interictal 30-45 Hz aperiodic exponent did not identify seizure onset contacts, but delta power and aperiodic offset did (bipolar montage, 5-min clips) sci:sha256:6c681383f1e1d4555a39ee5579bb262f7416fa29967d7dfc67c366d8cb549035

    Pipeline D: per-patient SOZ AUC table copied unchanged and compared with the other five pipelines.

  11. The 30–45 Hz aperiodic exponent did not separate seizure-onset channels in 31 SEEG patients sci:sha256:970015c7ad657e77ac8f470146719405ea5536dba9b4dabf2cdf20f1f8c23615

    [Legacy relation: reproduces] Per-patient amplitude-feature AUCs of pipeline F agree with the other pipelines (ICC about 0.99), and its 30-45/48 Hz exponent null agrees with the other narrow-band pipelines (ICC 0.88-0.91); the exponent null is band-specific.

  12. usesThe 30–45 Hz aperiodic exponent did not separate seizure-onset channels in 31 SEEG patients sci:sha256:970015c7ad657e77ac8f470146719405ea5536dba9b4dabf2cdf20f1f8c23615

    Pipeline F: per-patient SOZ AUC table copied unchanged and compared with the other five pipelines.

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

    The six reported AUC tables were processed from HUP iEEG in OpenNeuro ds004100 v1.1.3; the dataset is the source of the analysed results.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.