Stimulated pair angle did not improve distance-based prediction of N1 CCEPs in eight SEEG patients

We tested whether the orientation of a bipolar stimulation pair predicts expert-checked N1 cortico-cortical evoked potentials (CCEPs) beyond distance. The analysis used 20,308 pair-contact observations from eight SEEG patients in OpenNeuro ds005448 v1.0.0; recording contacts were on a different shaft from the stimulated pair. At 1–2 mA, adding angle changed leave-one-stimulated-pair-out AUC by a median of −0.0003 (three of eight patients positive; exact two-sided sign-flip p=0.875). The dipole-field score f/d³ had a lower AUC than inverse distance 1/d in five of eight patients (median change −0.0019, p=0.250). The pooled Mantel–Haenszel odds ratio for axial versus transverse contacts was 0.99. Angle coefficients varied in direction across patients; the group coefficient was not different from zero (p=0.578). The observed median coefficient for log f was 0.09, below the dipole-field prediction median of 1.25 in six of eight patients (p=0.086). Gray-matter-only, distance-limited, all-SEEG, and exploratory ECoG analyses showed no significant angle effect. These results suggest that pair orientation is not a useful control for aiming 1–2 mA single pulses at far SEEG targets, while retaining uncertainty: only eight SEEG patients were studied, small effects cannot be excluded, and the analysis tested binary N1 labels rather than amplitudes.

cortico-cortical evoked potentials · SEEG · electrical stimulation · distance · dipole field · brain-computer interface