Key Takeaways
- Neurosurgeons at London’s National Hospital for Neurology and Neurosurgery performed the world’s first AI‑assisted brain‑tumour removal, preserving a patient’s sight.
- The AI system analysed live endoscopic video, colour‑coding critical structures (nerves, blood vessels, pituitary tissue) to guide the surgical team in real time.
- Patient Rhys Hibbert, 48, regained full vision within a week and returned to work as a customer‑service manager without glasses or walking aids.
- The technology, developed by UCL researchers, was trained on hundreds of surgical videos, giving it exposure far beyond what a single surgeon could acquire in a career.
- Although the surgeons remained in full control, the AI acted as a safety net, reducing the risk of millimetre‑scale errors that could cause blindness, stroke or death.
- The operation was part of a NIHR‑funded clinical trial, highlighting the potential of AI to augment human expertise in highly delicate neurosurgical procedures.
Patient Background and Clinical Need
Rhys Hibbert, a 48‑year‑old customer‑service manager from Bedfordshire, was diagnosed in 2024 with a pituitary‑gland tumour that caused severe hormonal imbalance and progressive visual deterioration. Initially managed conservatively, his symptoms worsened to the point where health officials warned that without intervention he faced imminent blindness. “I could see everything in the room clearly” after the surgery, Hibbert later recalled, underscoring how his vision had been compromised pre‑operatively. The tumour measured 11 mm (0.4 in) and lay in a densely packed region where the optic nerves, pituitary stalk, and surrounding vasculature intertwine, making any deviation of even a millimetre potentially catastrophic.
The AI System: From Research Tool to Intra‑operative Aid
The artificial‑intelligence platform employed in the procedure had previously served only as a research tool at the National Hospital for Neurology and Neurosurgery (NHNN). Developed by a team led by Dr Sophia Bano, associate professor in robotics and AI at University College London, the system was trained on hundreds of operative videos, exposing it to a breadth of surgical scenarios that would take a human surgeon many years to encounter. Dr Bano explained, “By learning from hundreds of surgical videos, [the system] has been exposed to a breadth of surgical examples that would take a surgeon many years to encounter. It is designed to help recognise critical anatomy, surgical instruments and tissue interactions in real time, supporting the surgeon during highly delicate procedures.” During the operation, the AI processed live camera feeds, colour‑coding nerves, blood vessels, and tumour margins to highlight structures that must be avoided.
Surgical Workflow: Surgeon‑Led, AI‑Enhanced
Although the AI provided real‑time visual cues, the surgical team retained full control throughout the procedure. Surgeons used the colour‑coded overlay as an adjunct to their microscope view, allowing them to differentiate subtle tissue layers that are otherwise difficult to discern. The pituitary gland, optic nerves, and surrounding microvasculature occupy a space measured in millimetres; “going a millimetre wrong can make a critical difference,” health officials warned, noting that errors could result in blindness, stroke, or death. By integrating the AI’s annotations, the team could navigate this tight anatomical corridor with heightened confidence, ultimately excising the tumour while preserving visual pathways.
Immediate Post‑operative Outcome
Within days of the May surgery, Hibbert reported a dramatic improvement in his visual field. He stated, “When I came round … I could see everything in the room clearly,” and noted that he was able to walk independently without glasses or sticks within a week. His rapid recovery allowed him to resume his role as a customer‑service manager, describing his perception as “like I’ve got a 360‑degree panoramic view of everything around me.” The preservation of sight was the primary goal, given the tumour’s proximity to the optic apparatus, and the outcome met—and exceeded—expectations.
Technical Validation and Training Data
The AI’s performance relied on a robust training dataset comprising de‑identified surgical videos from prior pituitary and neurosurgical cases. This dataset enabled the model to learn the visual signatures of critical structures under varying lighting, angles, and tissue conditions. Dr Bano emphasized that the system’s real‑time recognition of instruments and tissue interactions was not a static lookup but a dynamic inference process that adapts to the operative field as it evolves. The model’s output was displayed on the surgeon’s monitor as a semi‑transparent overlay, ensuring that the AI’s suggestions complemented rather than obstructed the surgeon’s direct view.
Broader Implications for Neurosurgery
Prof Mike Lewis, scientific director for innovation at the National Institute for Health and Care Research (NIHR), which funded the operation as part of a clinical trial, lauded the procedure as a “pioneering surgery” that demonstrates the potential of advanced AI to support surgeons and improve patient care. He noted that such technology could reduce the learning curve for complex neurosurgical techniques, decrease intra‑operative complications, and expand access to high‑precision care in centres lacking subspecialty expertise. The NHNN’s historic legacy—founded in 1859 as the world’s first dedicated neurosurgical hospital—adds symbolic weight to the achievement, as Hibbert himself remarked: “The NHNN was founded in 1859 … and was the world’s first dedicated neurosurgical hospital. So to me, it seems very fitting that the same hospital should also be the world’s first to carry out an AI‑assisted neurosurgery.”
Patient Perspective and Future Outlook
Hibbert’s experience underscores the human impact of technological innovation in medicine. Beyond the clinical success, his return to normal life—free of visual aids and able to resume professional duties—illustrates how AI‑assisted surgery can translate into tangible quality‑of‑life gains. He expressed optimism about the future, suggesting that similar systems could become routine for other skull‑base tumours, vascular lesions, or functional neurosurgery where anatomical precision is paramount. The surgical team plans to publish detailed operative data and AI performance metrics, contributing to the evidence base needed for wider adoption and regulatory approval.
Conclusion: A Milestone in AI‑Enhanced Surgery
The world’s first AI‑assisted brain‑tumour resection at NHNN represents a landmark convergence of surgical expertise and machine intelligence. By providing real‑time, colour‑coded guidance of vital nerves and vasculature, the AI system acted as a safety net that helped preserve Hibbert’s sight while enabling a complete tumour excision. The operation’s success, coupled with the patient’s rapid functional recovery and return to work, highlights the promise of AI to augment human capability in the most delicate of medical interventions. As further trials validate and refine these tools, the neurosurgical community may witness a new era where technology and tradition collaborate to push the boundaries of what is surgically achievable.
https://www.theguardian.com/technology/2026/aug/27/london-neurosurgeons-ai-assisted-operation-brain-tumour