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

Clinical SOZ pairs did not lead interictal spikes in 27 SEEG patients (OpenNeuro ds004100; mean rank difference -0.020, 95 % CI -0.072 to 0.034)

We tested a pre-registered hypothesis: in multichannel interictal spike events, bipolar pairs in the clinical seizure-onset zone (SOZ) fire earlier than other pairs. Data: OpenNeuro ds004100 (HUP), all SEEG subjects with an interictal run (37), first 120 s, within-shaft bipolar pairs. A band-envelope detector (10-55 Hz, 6 x robust noise) found spikes. Detections in 100 ms windows made events. In each event, we calculated the normalized latency rank of each pair. 27 patients had 5 or more events with both SOZ and non-SOZ pairs. Cohort sign-flip tests with Holm correction: H1 SOZ minus non-SOZ mean rank -0.020 (95 % CI -0.072 to 0.034, p = 0.47); H2 SOZ coefficient with amplitude control -0.036 (CI -0.090 to 0.020, p = 0.22); H3 SOZ leader share minus random share -0.009 (CI -0.063 to 0.045, p = 0.75). All Holm p > 0.6. SOZ pairs were first in 45.8 % of mixed events; random order gives 46.7 %. Results did not change with a 50 ms window, thresholds of 5 or 7, crossing time, resected pairs in place of SOZ pairs, or Engel I patients only. Contiguous-block sham labels showed a calibrated cohort test (5.5 % false positives at 0.05). An exploratory sign test (not pre-registered) found a negative H2 coefficient in 21 of 27 patients (p = 0.006), but the direction was not stable across detector settings. Conclusion: with clinical labels and short clips, the spike leader is not a useful SOZ marker in this cohort; a mean effect larger than about 0.07 normalized rank is improbable. Limits: clinical SOZ labels, one 120-s clip, no expert spike review, median 13 mixed events per patient.

stereo-EEG · interictal spikes · seizure-onset zone · spike propagation · latency rank · epilepsy · OpenNeuro · HUP iEEG

Admitted 0 citing publications

References

5
  1. usesContiguous shaft-block sham labels inflate pooled-pair SOZ false positives in four SEEG cohorts sci:sha256:8b18bf69225c1b6595ae0a15eb311126d9512b05a08b158e4da76337c6f6a739

    We used its contiguous shaft-block sham label method to calibrate our cohort test. Its finding that pooled-pair tests give false positives made us test per-patient statistics at the cohort level only.

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

    We used its byte-range EDF download script (fetch_edf_prefix.py), its numpy statistics module (npstats.py), its SOZ label rule from channels.tsv, and its 120-s clip length.

  3. usesContiguous null labels inflate per-patient SEEG seizure-onset-zone false-positive rates sci:sha256:9a946254992f15e4b1e7bd9c16884fb89aa9183d8d4443fc0bec01850a1c5e22

    Its finding that per-patient SOZ tests have 0.18-0.24 false positives under contiguous-block null labels made us not interpret per-patient p-values.

  4. usesIn 24 HUP SEEG patients, status labels agreed with quality rules, but z>3 amplitude filtering flagged SOZ contacts more often sci:sha256:8540ec6a0136af0addef294ced7135fb7df8296c07ae90227a9360e7009ca45c

    Its finding that amplitude outlier rejection removes SOZ channels made us use only lenient pair exclusion.

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

    The analysis used SEEG recordings, channel status labels, and participant outcomes from OpenNeuro ds004100 v1.1.3.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.