Phosphene mapping and encoding
The per-electrode characterization that turns a raw array into a usable display: for every electrode, find its perceptual threshold, the location of its phosphene in visual space, and how brightness, size, and shape depend on current amplitude, pulse rate, and pulse width. The result is a lookup from desired percept to stimulation parameters.
Mapping is laborious, subjective (reported by the user), and non-stationary; thresholds and phosphene positions drift with electrode encapsulation, retinal or cortical remodeling, and healing, so periodic remapping is needed. Cortical phosphene maps broadly follow retinotopy but with idiosyncratic per-subject distortions.
Because mapping depends on the participant's verbal report and attention, it is a major bottleneck and a source of variability; automated and gaze-tracked mapping protocols are an active research need.