Within-patient SEEG sites under 10 mm apart showed excess percept-quality agreement, but effects varied across patients

We tested whether nearby stereo-electroencephalography (SEEG) stimulation sites in the same patient and hemisphere evoked visual percepts with more similar qualities. We used percept labels and native-space contact coordinates from OpenNeuro ds005169 v1.0.0 (43–50 Hz direct cortical stimulation; 20 patients). Each site was the midpoint of its two stimulated contacts. For 420 pairs among 104 sites, agreement was the fraction of three binary quality labels (dynamic, intermittent hallucination, color) that matched. A 10,000-permutation null shuffled label vectors within patient and hemisphere. Quality agreement decreased with distance (Spearman rho = -0.187, one-sided permutation p = 0.027). Pairs 0–10 mm apart agreed more than the null (0.762 vs. 0.694; p = 0.023); pairs at 10–20, 20–40, and >40 mm were not above the null. Results remained after adjustment for current difference (p = 0.028) and session gap (p = 0.042), and when seven missing contact coordinates were imputed along their shafts (rho = -0.223, p = 0.011). Same-shaft pairs at 20 mm or less had 0.147 greater agreement than other pairs (p = 0.042). Location-label agreement had a stronger distance association (rho = -0.233, p = 0.0002), with excess agreement in the 0–10 and 10–20 mm bins. No individual quality label was significant after Holm correction, and only six of ten patient-hemisphere groups with at least four sites had negative rho (sign test p = 0.25); the pooled effect was not consistent across patients. Each session contained one stimulation, so the planned same-session control was replaced by adjustment for session-index gap. The small sample, binary patient reports, native-space coordinates, possible regional confounding, and pooled result driven by patients with many sites limit interpretation. The result suggests that contacts more than approximately 10 mm apart may evoke independent percept qualities, but does not establish a consistent effect in each patient.

stereo-electroencephalography · SEEG · direct cortical stimulation · visual percepts · percept quality · site distance · permutation test · visual prosthesis