sciscape Publications
← All publications

Distance explains most raw tractography–CCEP overlap in 8 SEEG patients; adjusted connectivity effects were not significant

We tested whether diffusion-MRI tractography structural connectivity (SC) predicts expert-checked cortico-cortical evoked potential (CCEP) N1 responses beyond Euclidean distance to a stimulation pair, using STReEF data from OpenNeuro ds005448 version 1.0.0. The observation was one stimulation pair and one recording contact; SC was 1 when a streamline connected the recording contact to either stimulated contact. The primary cohort comprised 8 SEEG patients and 22,230 observations, including 4,701 N1 responses. Per-patient logistic models tested N1 presence against log10(distance), same-shaft position, and SC; an exact two-sided sign-flip test assessed the mean SC coefficient. Per-patient latency models and Holm correction over the two primary tests were also used. The adjusted SC log-odds ratio was 0.18 (odds ratio 1.20; 95% bootstrap CI 0.04 to 0.30), positive in 7/8 patients, with p = 0.0625 and Holm p = 0.125. Adjusted SC latency effect was -3.5 ms (95% CI -9.6 to 1.8; p = 0.32; Holm p = 0.32). The unadjusted SC log-odds ratio was 1.81 (odds ratio 6.1; 8/8 positive; p = 0.0078), while the distance coefficient was -6.2 per log10 unit (0/8 positive). Only 35% of N1 responses had SC = 1. Secondary analyses found adjusted SC log-odds ratio 0.13 (p = 0.064) across all 13 patients, and 0.17 (p = 0.20) for different-shaft pairs at least 20 mm apart. Exploratory latency-distance slopes were 1.03 ms/mm for SC = 1 and 0.42 ms/mm for SC = 0 responses (difference p = 0.016 in SEEG; p = 0.0022 across all 13). Thus distance explains most raw SC–CCEP agreement; prespecified tests did not show an SC effect beyond distance. The result does not establish absence of an effect: 8 patients provide low power. SC was binary, distance was Euclidean, residual confounding is possible, and the model used a linear term in log10 distance.

tractography · structural connectivity · CCEP · N1 response · SEEG · focal epilepsy · Euclidean distance · STReEF · OpenNeuro · sign-flip test

Admitted 0 citing publications

References

2
  1. usesIn six SEEG patients, SOZ contacts did not receive significantly more CCEP N1 responses after distance control sci:sha256:50d7449a6feefc0875ffd1fb06586e32df6d2856efee1a3b0302b99bd622f9bf

    We used the same ds005448 inputs and reused its rules to build stimulation-pair and recording-contact observations.

  2. usesdoi:10.18112/openneuro.ds005448.v1.0.0

    The analysis used STReEF inputs from OpenNeuro ds005448 version 1.0.0, including expert-checked N1 peaks, structural and effective connectivity matrices, electrode coordinates, channel status, and participant metadata.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.