SEEG site position did not predict color or dynamic visual percepts within patients

We tested whether the position of stimulated SEEG contact pairs within a patient predicted color percepts at lower sites or dynamic percepts at more lateral sites. The analysis used 112 stimulations from 20 patients in OpenNeuro ds005169 v1.0.0; all stimulations caused a percept, and no iEEG signals were used. Within-patient AUCs were 0.56 for color at lower z (90 events, 11 patients; permutation p = 0.25, Holm p = 0.49) and 0.43 for dynamic percepts at larger |x| (79 events, 9 patients; p = 0.79, Holm p = 0.79). Controls for current and imputed contacts did not change the conclusion. None of 18 exploratory tests passed FDR (smallest q = 0.25). The test sensitivity was limited to AUCs above approximately 0.63–0.64, so small or moderate effects remain possible. Native coordinates, coarse clinical labels, and few patients with both classes limit inference. These data do not support choosing contacts for color or motion from coordinates alone.

SEEG · cortical stimulation · visual percepts · color · dynamic perception · within-patient analysis · visual prosthesis