TL;DR

Surgeons at London’s NHNN have carried out what officials describe as the first successful AI-assisted removal of a brain tumour. The system read the operating camera feed live and colour-coded the structures the team needed to avoid. The patient kept his sight and was back at work within weeks.

Millimetres

Rhys Hibbert, 48, had an 11mm tumour on his pituitary gland. Diagnosed in 2024 and managed without surgery at first, his hormone levels and vision both deteriorated, and without an operation he faced losing his sight altogether.

That region is unforgiving. The gland sits packed against the vessels and the nerves that carry vision, and officials put it bluntly: a millimetre in the wrong direction can mean blindness, a stroke or death. The AI marked those structures on the surgeons’ view as they worked, while the team retained control throughout.

The operation took place in May at NHNN — the National Hospital, which specialises in neurology and neurosurgery, and is run by UCLH. Details were held back from publication until Hibbert had recovered. He walked unaided, without glasses or sticks, inside a week. “It feels like I’ve got a 360-degree panoramic view of everything around me,” he said.

Trained on hundreds of operations

Sophia Bano, the UCL robotics academic who led the technical work, said the system had been shown hundreds of recorded procedures — a range of surgical experience that would take an individual many years at the operating table to accumulate. The hospital had used the technology for research before, but never on a patient.

The NIHR, the government’s health research funder, paid for the procedure as part of a trial.

Looking forward

This is the fourth NHS AI diagnostic or clinical deployment Resultsense has covered this week, after chest X-ray triage in East Grinstead, a cleared physiotherapy pathway and blood clot scanning in Chelsea. The pattern is consistent: narrow tasks, human decision-makers kept in the loop, and outcome data attached.

It is also a reminder of where British AI adoption is actually working. While the labour market arguments dominate the headlines, trusts are quietly shipping systems that do one thing well. A single trial patient proves rather little on its own, and the honest read is that this is a promising start under research conditions rather than a service change.