World’s First Patient to Undergo AI-Assisted Brain Surgery Recovers Vision
Constantvpn.com – The world’s first patient to undergo a brain operation guided in real time by an artificial-intelligence system has had his tumour fully removed. Rhys Hibbert, a 48-year-old customer-services manager and father of two from Bedfordshire, faced potential blindness before surgeons at the National Hospital for Neurology and Neurosurgery used a bespoke AI tool to navigate around hidden vessels and nerves during his procedure.
How the Live AI Tool Works During Pituitary Surgery
Hibbert’s 11 mm non-cancerous growth sat on the pituitary gland, a marble-sized structure at the base of the skull that regulates hormones governing growth, metabolism, blood pressure, and body temperature. Because the gland nestles directly beside the carotid arteries and the optic nerves, even a small tumour can compress vision pathways. Hibbert’s growth had been squeezing his optic nerves, steadily narrowing his peripheral sight over roughly 18 months.
The operation reached the tumour through an endoscope—a thin camera on a flexible shaft—inserted via the nose to the skull base. At that depth the growth was close yet concealed behind bone and tough membrane, and individual anatomical variation made pinpointing the safest extraction route far from straightforward. Published complication data suggest a 25-to-50 percent chance of leaving residual tumour behind and a 0.5-to-2 percent risk of injuring a major vessel.
On a second monitor, the AI system analysed the live video feed, tracked the surgical instruments frame by frame, and overlaid probability maps showing where critical vessels and nerves were most likely to lie. The technology functions much like facial-recognition software on a smartphone, except it identifies deep anatomical structures rather than faces. Researchers trained and validated the model on hundreds of recorded pituitary-tumour removals, manually tracing vessels and nerves in each clip so the algorithm could learn to “recognise” the most consequential structures.
“It is trained on more operations than most surgeons see or perform in a lifetime and can act like an expert second pair of eyes,” explained Prof Hani Marcus, who co-led the procedure alongside Mr Danyal Khan.
Unlike earlier experimental systems, the tool operates in real time during surgery. The team stressed that the AI merely assists; the surgeons retain complete operative control. Their longer-term goal is to build a conversational, ChatGPT-style advisory layer that gives the operating team “another expert in the room they can turn to for advice if they want it—or leave if they disagree with it,” Prof Marcus said. The project is funded by the National Institute for Health and Care Research and Google, and the team has spent years validating the technology in the laboratory before moving to live surgery. A larger clinical trial is now under development.
Hibbert, who had been fit and walking two miles at lunchtime until he collapsed unconscious on a pavement and began fitting, said he did not hesitate when offered the chance to be the world’s first patient to undergo this AI-assisted approach. “If patients are not prepared to join research, how can doctors ever learn and how can medicine ever progress?” he asked. Eight weeks after the operation he reported his sight sharpening every few days, his energy returning, and all pre-operative side effects disappearing. “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 recalled.
Frequently Asked Questions
What exactly did the AI do during the surgery?
The system processed the live endoscopic video in real time, tracked instrument positions, and generated probability overlays indicating where major blood vessels and nerves were most likely located. It did not move any instruments or make operative decisions; surgeons retained full control at all times.
Was the tumour fully removed?
Surgeons confirmed the tumour was removed successfully. Hibbert reported progressive visual improvement in the weeks following the procedure and stated that pre-operative fatigue and dizziness had resolved.
How was the AI trained?
Researchers annotated hundreds of recorded pituitary-tumour removal videos, drawing around vessels and nerves in each clip so the model could learn to identify those structures. The system was then evaluated before being cleared for live intra-operative use.
Is this technology available to other hospitals yet?
The team is preparing a larger clinical trial. The current procedure was a single-case proof-of-concept at the National Hospital for Neurology and Neurosurgery, funded by the NIHR and Google, and wider availability depends on the outcomes of that trial.
