In two ds004977 SEEG sites, inter-pulse coupling added little predictive value for CCEP strength beyond distance

This pilot asked whether spontaneous coupling between a stimulation site and a recording site, measured between stimulation pulses, predicted single-pulse cortico-cortical evoked potential (CCEP) strength beyond Euclidean distance. It used two sites from OpenNeuro ds004977 v1.2.0: sub-1 ROC11-ROC12 (179 recording channels) and sub-4 RO3-RO4 (98 channels). In this snapshot, each analyzed run contained one contiguous uncensored span of about 53 s with one stimulation site and 12 pulses; only six inter-pulse windows (about 36 s total) contributed coupling estimates per site. The analysis compared 12 coupling measures with CCEP log SNR using distance-adjusted partial Spearman correlation, permutation tests within distance quantile bins, and 10-fold cross-validated R2. Distance-only CV R2 was 0.135 for sub-1 and 0.742 for sub-4. Adding coupling changed CV R2 by -0.013 to +0.013 across all 24 site-by-measure tests. Two tests had uncorrected p < 0.05 at sub-4, but neither replicated at sub-1 and neither survived Bonferroni correction. This two-site pilot found little added predictive value from inter-pulse coupling beyond distance. The short coupling windows make estimates noisy, so this result does not exclude an effect with longer baseline recordings. Coupling was measured during stimulation, not at separate rest, and late CCEP tails may remain.

SEEG · cortico-cortical evoked potentials · inter-pulse coupling · coherency · imaginary coherency · distance · OpenNeuro · pilot study