Pierwsza operacja neurochirurgiczna z bezpośrednim udziałem sztucznej inteligencji
- First live AI-assisted neurosurgery: An artificial intelligence system was used for the first time to assist a neurosurgeon in real time during a brain tumor removal at the National Hospital for Neurology and Neurosurgery (UCLH).
- Core functionality: Developed at University College London (UCL), the AI analyzed live endoscopic video feeds during the operation, highlighting critical anatomical structures to prevent complications such as stroke, vision loss, or fatal injury.
- Patient background:
- Rhys Hibbert, a 46-year-old patient from Bedfordshire, was diagnosed with an 11 mm pituitary gland tumor in 2024 after collapsing and suffering a seizure.
- Due to worsening hormonal issues, vision impairment, and reduced mobility, he consented to participate in the clinical trial as the first patient operated on using this system.
- Advantages over traditional planning:
- Unlike static pre-operative scans, live surgery involves shifted tissues, shifting camera angles, blood, and obstruction by surgical instruments.
- The intraoperative AI adapts dynamically to what the surgeon is actively seeing.
- Training and future roadmap:
- The algorithm was trained on hundreds of historical pituitary surgery videos, encompassing scenarios equivalent to years of clinical experience.
- Next iterations aim to track surgical tools and their interactions with tissue to provide guidance during complex maneuvers.
- Outcome: The procedure was a success; the patient reported immediate visual improvement upon waking and was walking unassisted without glasses within a week.