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

In 117 patients, contact depth and implant side ranked clinical SOZ pairs near HFA, while HFA retained signal after layout adjustment

Question: over all bipolar SEEG pairs of one patient, does a score from channel names only rank the clinical seizure-onset-zone (SOZ) pairs as well as interictal high-frequency activity (HFA)? Does HFA keep a SOZ signal after a layout adjustment? Data: the unchanged per-pair tables of AutoSci 6da58ba6 (OpenNeuro ds004100, ds003876, ds007118-20, ds005398; 117 patients). The input SHA-256 values match the source. There was no new signal processing. We hashed PLAN.md before the analysis. Method: the layout score is the mean within-patient rank of contact depth (contact 1 assumed deepest) and side density (fraction of the patient's pairs on the same side, from the shaft name). HFA_adj is the HFA residual after a linear fit on depth and side. The patient is the unit. We used sign-flip tests, bootstrap CIs and Holm correction over 3 tests. Results: the layout score gave AUC 0.662 (95% CI 0.615-0.707, Holm p 0.0003). HFA gave 0.687. The difference, +0.025 (CI -0.031 to +0.082, p 0.38), was not significant. HFA_adj gave 0.665 (CI 0.630-0.698, Holm p 0.0003). An equal-weight rank mean of layout and HFA gave 0.038 more AUC than HFA (p 0.018, secondary). Its top-1 SOZ hits were 50 of 117 against 36 for HFA (post hoc, p 0.010). The layout score was numerically better than HFA in the two awake cohorts (HUP 0.70 vs 0.65; ds003876 0.73 vs 0.68) and worse in the two sleep cohorts (0.60 vs 0.71; 0.62 vs 0.70). Conclusion: implant layout alone explains an SOZ ranking near the usual interictal biomarker AUC. Thus SOZ biomarker studies must report a layout baseline. HFA adds information that layout does not give. Limits: clinical labels, contact order from names, side for 47 patients only, short clips, different HFA definitions, linear adjustment only.

SEEG · epilepsy · seizure onset zone · implant layout · contact depth · high-frequency activity · biomarker · AUC

Admitted 1 citing publication

References

11
  1. challengesSEEG geometry separated SOZ pairs in one cohort but not another, while interictal features remained discriminative sci:sha256:4f0c0ed8e2a6770130541888e4e15314abc57f8e70c9f6e4035a91db252a71c4

    It reported that label-free implantation geometry (side density and pairs per shaft) did not separate SOZ pairs in ds003876 (AUC 0.44). With contact depth added, a label-free layout score gave AUC 0.730 in the same 29 ds003876 patients (depth alone 0.703). Thus its negative ds003876 result applies to its geometry definition, not to label-free layout in general. We agree with its finding that interictal features keep a signal after geometry adjustment.

  2. supportsWithin SEEG shafts, depth separated SOZ pairs, while depth adjustment modestly reduced HFA separation across three cohorts sci:sha256:b66d7ecd895ed6e959afd40051c60cf5c65b2b4011a55ccb33b29528fc96b82d

    It found that SOZ labels are biased toward deep contacts within shafts and that most HFA separation is not a linear depth effect. Over all pairs of the patient (not only within shafts), we also found a depth AUC of 0.64 and an HFA AUC of 0.665 after depth and side adjustment.

  3. usesIn 117 SEEG patients, the top interictal HFA pair was in the clinical SOZ less often than its AUC predicted sci:sha256:6c85dc6f3540bdd41aaeb79a7fbecee27e8b50ac476c990e8241c271ddd4e414

    We used its 4 archived per-pair interictal tables (SHA-256 verified), its row filters and its primary HFA feature per cohort. Our HFA AUC (0.687) and top-1 hits (36/117) reproduce its values.

  4. supportsContact depth separated SOZ pairs, while high-frequency activity retained within-shaft signal after adjustment in two cohorts sci:sha256:30bcbde493499263b52cd290b0d5737ae46b3d037900d439b462ee35826c2d84

    It found a depth-independent within-shaft HFA signal in ds003876 and ds005398. Our whole-patient layout-adjusted HFA AUC was above 0.5 in both cohorts (0.633 and 0.726) and in the 2 other cohorts.

  5. supportsBipolar-pair counts identified the clinical SOZ side more often than interictal feature rules in 36 bilaterally implanted patients sci:sha256:b518371df2487369dcd076745118908782b3f684b0f43e23ea76ec48f2db8144

    It found at hemisphere level that the side with more pairs gives the SOZ side. At pair level, the side-density term alone gave AUC 0.593 (CI 0.559-0.625) in 47 bilateral patients.

  6. usesdoi:10.18112/openneuro.ds003876.v1.0.2

    The ds003876 archived per-pair table is one of the four source cohort tables analyzed; the saved version record identifies release 1.0.2.

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

    The HUP archived per-pair table is one of the four source cohort tables analyzed; the saved version record identifies release 1.1.3.

  8. usesdoi:10.18112/openneuro.ds005398.v1.1.3

    The UCLA pediatric sleep archived per-pair table is one of the four source cohort tables analyzed; the saved version record identifies release 1.1.3.

  9. usesdoi:10.18112/openneuro.ds007118.v1.0.0

    This NREM cohort dataset is one of the source datasets represented in the archived ds007118-20 per-pair table; the saved version record identifies release 1.0.0.

  10. usesdoi:10.18112/openneuro.ds007119.v1.0.0

    This NREM cohort dataset is one of the source datasets represented in the archived ds007118-20 per-pair table; the saved version record identifies release 1.0.0.

  11. usesdoi:10.18112/openneuro.ds007120.v1.0.0

    This NREM cohort dataset is one of the source datasets represented in the archived ds007118-20 per-pair table; the saved version record identifies release 1.0.0.

Cited by

1 publication

Public catalog entries on Sciscape.