Monash Vision Group is developing a vision prosthesis based on implanting stimulating tiles into the visual cortex, area V1. Each tile has 43 stimulating electrodes that are designed to stimulate Layer 4 of V1 using bi-phasic current pulses. The tiles wirelessly receive coded messages from an external transmitter, which itself receives data from a vision processor. The vision processor extracts important scenic features from a small camera mounted in the headgear, using several algorithms designed for common everyday tasks, such as navigation and social/business interactions.  A complete end-to-end system has been developed and is in the pre-clinical phase of testing.

This talk will first cover the engineering development of the project, which includes the design, development and testing of: insulated electrodes with ablated annulus stimulation regions, a 500,000-transistor ASIC chip, a hermetic package, wireless link (transmitter and receiver), processor, and software algorithms.  Secondly, the complex trade-offs in designing a complete system will be discussed, including the basic physical/physiological principles that govern the ultimate performance of cortical interfaces.  Finally, futuristic versions of cortical interfaces that propose substantially greater numbers of stimulation points will be explored. Suggestions about the areas of research that could underpin such developments will be made.