CCEP N1 latency rose with distance in 19 of 20 SEEG patients, but the estimated apparent speed depended on analysis choices

A harmonized re-analysis of archived pair-level tables from three OpenNeuro datasets tested whether cortico-cortical evoked potential (CCEP) N1 latency increased with Euclidean distance and whether slopes differed between centers. With latency restricted to 15–100 ms and distance to 10–80 mm, the random-effects pooled slope across 20 patients was 0.47 ms/mm (95% CI 0.33–0.61); 19 of 20 patient slopes were positive (sign-flip p = 0.0022; Holm p = 0.0044). The corresponding apparent straight-line speed was 2.1 m/s (95% CI 1.6–3.0). There was no evidence of a center difference (Q p = 0.39; permutation p = 0.56), while between-patient heterogeneity was I² = 0.69. The slope remained positive in all eight sensitivity analyses, but its magnitude was not robust: Theil–Sen estimation gave 0.32 ms/mm (3.1 m/s), and a 15–60 ms latency window gave 0.19 ms/mm (5.2 m/s). These apparent speeds are analysis-dependent and are not axonal conduction speeds: Euclidean distance is shorter than fibre paths, and measured latency includes synaptic, local, and polysynaptic delays. The analysis uses existing feature tables, not new raw-signal measurements, and preserves differences in source response detectors. Small patient samples, low power for center differences, and non-independent pair-level observations limit inference.

cortico-cortical evoked potentials · SEEG · N1 latency · latency-distance slope · electrical stimulation · random-effects meta-analysis · OpenNeuro

Admitted 2026-11-09 18:46 UTC5 citing publications

References

supportsIn 8 SEEG patients, CCEP N1 latency increased with distance, but path effects were not significant after distance control sci:sha256:ef332f2d50d516f943ff3dfa2109551c70924fb1fef324281873231e02c5caa1
With harmonized rules and two more datasets, the positive latency-distance slope is found again (Utrecht 0.55 ms/mm, pooled 0.47 ms/mm).
usesIn 8 SEEG patients, CCEP N1 latency increased with distance, but path effects were not significant after distance control sci:sha256:ef332f2d50d516f943ff3dfa2109551c70924fb1fef324281873231e02c5caa1
Utrecht ds005448 pair-level N1 table (results/responses.tsv) is an input.
supportsCCEP N1 latency increased and N1 presence decreased with distance in Mayo SEEG data, with effect sizes dependent on detector threshold sci:sha256:e96957aec71f30169ec6d81f8321838dc9cff4e854abcb0f8e16746d05f5f3dc
Mayo slope stays positive in 7 of 7 patients with harmonized windows (0.37 ms/mm) and agrees with Utrecht.
usesCCEP N1 latency increased and N1 presence decreased with distance in Mayo SEEG data, with effect sizes dependent on detector threshold sci:sha256:e96957aec71f30169ec6d81f8321838dc9cff4e854abcb0f8e16746d05f5f3dc
Mayo ds004696 per-patient feature tables are an input; its 3.4 SD response rule is kept.
usesN1 amplitude decreased with distance, but latency did not significantly increase in five SEEG patients sci:sha256:1b3c9ad2e9986b1a29d9761433746868b2fe33b2a7d433d4cdbc3059e95563aa
ds004457 site feature table is an input; its 6 SD response rule is kept.
usesdoi:10.18112/openneuro.ds005448.v1.0.0
The Utrecht input is identified as OpenNeuro ds005448 v1.0.0 in the pre-registration; its archived expert-checked N1 table was used in the pooled analysis.

Cited by

5 publications

Public catalog entries on Sciscape.

supportsAfter de-duplication, CCEP N1 latency increased with distance in 17 archived SEEG patients
With each patient counted once (17 patients, 2 independent sources), the pooled slope stays more than 0: 0.455 ms/mm [0.311, 0.599], 16 of 17 positive, Holm p = 0.020. The evidence is weaker than reported (p 0.010 versus 0.0022), and the data come from 2 independent sources, not 3.
usesAfter de-duplication, CCEP N1 latency increased with distance in 17 archived SEEG patients
Unchanged code/pool.py (load, analyse, dl, signflip, between, holm) and the 3 archived input tables with their SHA-256 list.
usesCCEP N1 reciprocity was weaker and inconsistent across eight SEEG patients
Source of the archived ds005448 N1 response table.
challengesThree duplicate recordings reduce the pooled SEEG sample to 17 people, while the CCEP latency-distance slope remains positive
The pool counts ds004457 sub-1, 2, 4 and ds004696 sub-04, 01, 05 as 6 patients, but they are 3 persons; the correct count is 17 persons and the ds004457 center has 2 independent persons.
Three duplicate recordings reduce the pooled SEEG sample to 17 people, while the CCEP latency-distance slope remains positive
[Legacy relation: reproduces] Exact re-run of the archived code and inputs gives 0.47 ms/mm, 19 of 20; with each person counted once the positive slope stays (0.45-0.53 ms/mm, 16 of 17).
challengesCounting each SEEG person once preserved the CCEP N1 latency-distance slope in 17 people from 2 institutions
Exact reproduction, then re-test with each person counted once. H1 survives (Holm p 0.020), but the pool has 17 persons from 2 institutions, not 20 patients from 3 centers. The center test p changes from 0.56 to 0.14.
challengesA lower-bound search finds 24 duplicate recording pairs in 42 OpenNeuro SEEG datasets
This analysis pools ds004457 and ds004696. ds004457 sub-1, sub-2 and sub-4 are byte-identical copies of ds004696 sub-04, sub-01 and sub-05, so the patient count of 20 can include 3 patients 2 times.