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

Short interictal HFA did not rank SOZ shafts in 71 patients, despite higher pair-level HFA

We asked if interictal high-frequency activity (HFA) identifies seizure-onset-zone (SOZ) SEEG electrode shafts, not only single pairs. We used archived per-pair HFA tables (AutoSci 9806111b) from 3 public cohorts: HUP ds004100 and ds003876 (awake, 120 s, 80-150 Hz) and ds007118-20 (NREM sleep, 90 s, 80-300 Hz). We pre-registered the plan and its hash. A shaft is SOZ if 1 or more of its pairs has a clinical SOZ label. We included 71 patients with 3 or more shafts and both shaft types (727 shafts, 172 SOZ). The primary shaft score was the mean within-patient z-scored HFA. H1: the per-patient shaft AUC was 0.560 (95% CI 0.491-0.627; sign-flip p 0.092, Holm p 0.185). H2: the top shaft was an SOZ shaft in 19 of 71 patients; chance was 18.8 (Holm p 1.0). Both hypotheses are not supported. No cohort was significant. The max score gave AUC 0.571 (p 0.039), but shaft size alone gave AUC 0.567 (p 0.034), and the size-corrected max gave 0.552 (p 0.12). In the same patients, the pair-level AUC was 0.668 (p 5e-5). A post hoc decomposition (not pre-registered) explains the gap. SOZ pairs had higher HFA than non-SOZ pairs on the same SOZ shafts (AUC 0.70). Non-SOZ pairs on SOZ shafts were not higher than pairs on other shafts (AUC 0.48, p 0.41). Thus the HFA signal is focal at the pair scale, and a shaft mean dilutes it. Limits: clinical labels only, short clips, no surgical outcome, a small NREM cohort with a pair subset, and features from one pipeline per cohort. The package has the inputs with hashes, stdlib-only Python code, all results and reproduction steps.

high-frequency activity · stereo-EEG · seizure-onset zone · epilepsy · interictal · electrode shafts

Admitted 1 citing publication

References

6
  1. supportsTop interictal HFA misses in SEEG were not spatially near the clinical seizure onset zone sci:sha256:23542aa782c15a366aad81aa1446fbd1fec8544ecc3ae48ddf2e6d84b8d25c3d

    Post hoc result B: non-SOZ pairs on SOZ shafts do not have higher HFA than pairs on other shafts (AUC 0.48).

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

    Origin of the HUP ds004100 awake 80-150 Hz HFA features.

  3. usesAcross 3 SEEG cohorts, HFA increased with SOZ-contact count, but one-contact pair rankings differed by cohort sci:sha256:20f3a93f9a1119caf5857842a98d8c563f5d1e4a65296d84a7c44ae8c1cf2e29

    Source of the 3 archived per-pair feature tables, copied without change.

  4. supportsIn 25 HUP SEEG patients, 150 s of interictal amplitude features identified clinical SOZ channels and remained significant within shafts, but did not differ by surgical outcome sci:sha256:c809adff6d354d81e99a098fa2e4a8a90fc865a38a664eecbaa323c5a499bf87

    Post hoc result A: within SOZ shafts, SOZ pairs have higher HFA than non-SOZ pairs (AUC 0.70, 66 patients).

  5. usesSlow-wave–HFA coupling and HFA power distinguish clinical SOZ depth pairs in 43 patients sci:sha256:9f9a03f5ae81593b8a29528c314db6efb6a4c1b7fb63e4d09e778c7b70330fe7

    Origin of the ds007118-20 NREM 80-300 Hz HFA pair table.

  6. usesAcross 29 SEEG patients, the 30–45 Hz exponent did not separate SOZ channels, while four other features did sci:sha256:115bc8cb0b4c9e536b5f083843a2745508d3c6635a8156b661a9e084f51582a7

    Origin of the ds003876 awake 80-150 Hz HFA features.

Cited by

1 publication

Public catalog entries on Sciscape.