Prosthetic Vision Simulation in Fully and Partially Sighted Individuals

Abstract

This article examines performance and learning under conditions of simulated prosthetic vision, using the tasks of counting white squares on modified checker boards, and placing black checkers on these squares. Aim of our study was to determine how much environmental information people would be able to obtain from crude pictures without tactile feedback. For this, we used video camera images which were convolved by a dedicated computer program in real time to represent phosphene vision as future chip implants might be able to do. The resolution was equivalent to a vision of 20/2400. The program allowed us to vary a broad set of variables such as contrast, number of gray levels, number of phosphenes per column and row, etc. In our tests we used a matrix of 6×10 dots with Gaussian intensity profile. Test subjects varied in age, gender, and educational background.