AI-Guided Brain Implant Learns a Blind Patient’s Visual Cortex to Deliver Sharper Light Patterns With Less Current
Reported by Gadget Review1 source ↓
An AI-guided implant stimulation method matched target brain activity more accurately and used less current than comparison methods in a one-person study. The experiment produced light sensations, not functional vision.
Key points
- The AI selected stimulation settings using the participant’s recorded brain responses
- The trial involved one 27-year-old man and a temporary 96-channel implant designed to remain for six months
- The participant reported flashes, spots and limited shapes of light; everyday sight was not demonstrated
- Long-term safety, durability and usefulness remain unknown
Details
The proof-of-concept study, published in Neuron, involved researchers from UC Santa Barbara, ETH Zurich and Miguel Hernández University. A 27-year-old man who lost his sight after a traumatic brain injury took part in a feasibility trial in Spain.
The system recorded the participant’s brain activity before stimulation, then selected electrical settings intended to produce a target pattern of activity. Researchers report that its selections matched target activity more accurately and required less current than comparison methods. Recorded brain responses also predicted his reported visual sensations more reliably than stimulation settings alone.
The participant experienced phosphenes—flashes, spots and limited shapes of light. The study did not demonstrate reading, navigation, face recognition or other functional everyday vision. The temporary array had 96 channels and was designed to remain in place for six months; the study involved only one participant, leaving long-term safety and durability unresolved.
Written from coverage as of
Images
- Gadget Review