Surgeons Remove Brain Tumor Using Real-Time AI Guidance

First Patient Undergoes AI-Assisted Brain Surgery

Surgeons at University College London Hospital have successfully removed a brain tumor from a patient using live artificial intelligence assistance during the operation—a medical first. Rhys Hibbert, a 48-year-old customer services manager from Bedfordshire, volunteered to be the inaugural patient for the experimental procedure.

Hibbert had developed an 11mm non-cancerous tumor on his pituitary gland, a marble-sized organ at the base of the skull that regulates vital hormones. The tumor's location presented significant surgical challenges: it sat dangerously close to the carotid arteries supplying blood to the brain and the optic nerves controlling vision. As the growth pressed against his optic nerves, Hibbert experienced narrowing peripheral vision, severe fatigue, and dizziness.

How the AI System Works

The AI tool analyzed a live video feed from an endoscope inserted through Hibbert's nose to reach the skull base. On a second screen, the system tracked surgical instruments in real time and marked areas where vessels and nerves were most likely located, highlighting the safest zones for tumor removal.

The technology functions similarly to facial recognition software but identifies anatomical structures instead of faces. Researchers trained the system on hundreds of videos of pituitary tumor removals, manually annotating vessels and nerves in each recording to teach the AI to recognize critical structures.

"It is trained on more operations than most surgeons see or do in a lifetime and can act like an expert second pair of eyes," said Prof. Hani Marcus, who helped perform the procedure.

The average UK surgeon performs only 10 to 20 of these operations annually, typically relying on pre-surgery scans to understand patient anatomy. The AI system provides additional guidance during the procedure itself, when anatomy may differ from expectations.

Why It Matters

Real-time AI surgical assistance represents a significant advance over experimental tools that analyze procedures only after they're complete. The technology could reduce complication rates in delicate surgeries where hidden anatomical variations create risk—in this case, a 25-50% chance of incomplete tumor removal and up to 2% risk of major blood vessel injury. The research team's long-term goal is developing a "surgical ChatGPT" that functions as an on-demand expert consultant surgeons can consult or override as needed.

Patient Recovery and Next Steps

Hibbert reported immediate vision improvement after surgery. "When I opened my eyes after the operation it felt like I had got 360-degree vision—I hadn't seen that clearly for a year," he said. Eight weeks post-surgery, his sight continues improving and his pre-operative symptoms have resolved.

The research team, funded by the National Institute for Health and Care Research and Google, spent years testing the technology in laboratory settings before this first clinical application. They are now planning a larger trial.

Prof. Marcus emphasized that surgeons retain full control during AI-assisted procedures—the system provides guidance that surgical teams can accept or reject based on their judgment.

These details were first reported by BBC News.