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

SEEG depth-contact N1 latency did not scale with distance in 16 patients, while ECoG latency did

This independent test examined whether cortico-cortical evoked potential N1 latency increases with distance from the stimulated pair at SEEG depth contacts in 16 patients from OpenNeuro ds004080. It analyzed 2,438 automatically detected negative-peak N1 units. The mean per-patient distance slope was 0.11 ms/mm (95% CI -0.11 to 0.32; 11/16 positive; exact sign-flip p=0.31). The confidence interval contains both zero and the 0.25 ms/mm estimate from earlier work, so the result does not distinguish no slope from that prior estimate. Distance did not improve within-patient or held-out-patient latency prediction. A stimulation-site offset reduced median within-patient error by 7.7 ms in all 16 patients. In the same patients, ECoG contacts showed a slope of 0.39 ms/mm (95% CI 0.30 to 0.47; 16/16 positive; p=3e-5). The likely explanation for the depth-contact result is the negative-only detector: prior work found that 55-62% of expert-checked depth N1s were positive. These findings warn against using negative-only automatic N1 latencies at SEEG depth contacts as stimulation-model timing targets; labels were automatic and the proposed polarity explanation remains probable, not proven.

cortico-cortical evoked potentials · N1 latency · SEEG · ECoG · electrical stimulation · distance · neural response · negative-peak detection

Admitted 1 citing publication

References

7
  1. supportsN1 CCEP peak latency increased with distance in 6 of 8 SEEG patients, but distance did not predict latency in held-out patients sci:sha256:985b480b21553ddaacc1944d354e3c77964a53382622135f9c3841617ae82055

    Distance does not improve held-out-patient latency prediction at ds004080 depth contacts. The slope at depth contacts (0.11 ms/mm, CI -0.11 to 0.32) is not significant, but its CI contains 0.25 ms/mm. ECoG contacts in the same patients give 0.39 ms/mm.

  2. supportsIn eight patients, stimulation-site offsets improved prediction of intracranial N1 latency more than distance sci:sha256:aff929e6fda1ba3606faf0dff7ace45bd3bcad46b3df41dd447a4b512ede7906

    In a second dataset (ds004080, 16 patients, SEEG depth contacts), distance also does not predict held-out latency, and a stimulation-site offset also decreases the timing error (by 7.7 ms, 16 of 16 patients).

  3. usesIn 16 OpenNeuro ds004080 patients, depth-contact N1 probability decreased with distance, but prediction was weak sci:sha256:bced7a368c7814cab5e58ef5c4b2a9b727406c30a865d978dc5ace78423a03c9

    This work detected the N1s from the raw ds004080 BrainVision files. Its negative-peak rule defines the N1 labels and latencies that are tested here.

  4. supportsIn eight SEEG patients, N1 latency rose with distance but geometry poorly predicted held-out patients sci:sha256:0ee0d8861516edeb49aa9ad1828f50017c23f4e71951bf0cc00119390431cde4

    That work found that 55-62% of expert depth N1s are positive. The near-uniform latencies of negative-only automatic N1s at ds004080 depth contacts agree with the conclusion that a negative-only rule is not valid at depth contacts.

  5. usesN1-map overlap decreased with stimulation-site separation in four patients at a second SEEG center sci:sha256:d5a7c95d283aff8a628675a2589d25da74a3025faf18beb02218094aa04c3264

    The 16 per-patient ds004080 N1 row files come from this package without change.

  6. usesdoi:10.18112/openneuro.ds004080.v1.2.4

    The analyzed per-patient rows were derived from OpenNeuro ds004080; the recorded exact-version metadata identifies v1.2.4 and its CC0 license.

  7. doi:10.1038/s41593-023-01272-0

Cited by

1 publication

Public catalog entries on Sciscape.

  • challengesN1 latency rises with distance at SEEG and ECoG contacts in one dataset

    64345c41 found no latency-distance slope at SEEG depth contacts in ds004080 (0.11 ms/mm), but 0.39 ms/mm at ECoG contacts. In ds005448, with one pipeline and expert-checked N1s, the depth slope (0.49 ms/mm, 7/8) is equal to the ECoG slope (0.49 ms/mm; difference p = 0.47). Thus the smaller depth slope is not a general depth property. This agrees with the detector explanation that 64345c41 gives.