An international team of scientists from UC Santa Barbara, ETH Zurich, and Miguel Hernández University has developed an AI-powered visual prosthesis system that dynamically adapts electrical brain stimulation to restore clearer visual perceptions in people with blindness. Testing the approach with a blind participant fitted with a visual cortex implant, researchers discovered that monitoring real-time resting neural activity allows deep-learning algorithms to predict the exact shape, size, and brightness of perceived light flashes (phosphenes). Rather than relying on rigid stimulation patterns, the personalized system continuously adjusts to an individual’s fluctuating brain state.
About the Author
