Particle.news

UCL Surgeons Use Real‑Time AI to Remove Pituitary Tumour

Colour‑coding of nerves and blood vessels during a May operation helped protect the patient’s sight, prompting plans for a larger clinical trial.

Overview

  • The operation, which took place in May, removed an 11mm pituitary tumour from 48‑year‑old Rhys Hibbert and restored his vision so he could return to work.
  • A UCL‑developed computer‑vision system analysed the live endoscopic video feed and colour‑coded critical anatomy such as optic nerves and blood vessels to guide the surgical team.
  • Surgeons led by Prof. Hani Marcus and resident Danyal Khan retained full control throughout the procedure and used the AI as a decision‑support tool rather than an autonomous actor.
  • The AI was trained on hundreds of annotated pituitary surgery videos at UCL’s Hawkes Institute and ran on NVIDIA’s Clara IGX platform with funding from the NIHR and Google.
  • UCL says the pilot’s success has led to a larger trial and further technical development, including instrument tracking and wider operating‑room integration, to test safety and effectiveness before broader use.