In 8 SEEG patients, distance-adjusted CCEP N1 out-degree did not predict contact in-degree

Question: do SEEG stimulation pairs that cause N1 responses at many contacts (high out-degree) also receive N1 responses from many pairs (high in-degree)? Data: expert-checked N1 detections, electrodes and channels tables of OpenNeuro ds005448 v1.0.0 (8 SEEG and 5 ECoG patients). Method (plan written before the statistics): per patient, a logistic model of N1 on log10 distance and same shaft gives residuals. Out-degree and in-degree are mean residuals. Spearman rho relates the out-degree of each pair to the mean in-degree of its two contacts. A shaft-split control removes all direct reciprocal edges. An exact sign-flip test across patients on Fisher z, with Holm correction, gives the group result. Result: the unadjusted rho is positive (11/13 patients, p = 0.0078; SEEG mean 0.165). After distance adjustment, the SEEG mean rho is -0.094 (95% CI -0.224 to 0.039, 2/8 positive, Holm p = 0.156). The shaft-split rho is -0.070 (95% CI -0.145 to 0.005, Holm p = 0.156). A tissue-adjusted model and a 13 mm minimum distance give the same conclusion. Thus the raw source-sink coupling comes from implant geometry, and a strong positive node-level hub trait is improbable. Limits: 8 SEEG patients, binary N1, simple distance model. The package has the code (including a numpy-only MATLAB v5 reader), the input files with SHA-256, the results and the commands to reproduce them.

CCEP · N1 response · SEEG · source-sink coupling · brain connectivity · electrode geometry