Surgeons in London claimed success in performing the world’s first AI-assisted surgery to remove a tumour in the brain that saved the vision of a 48-year-old man.
The operation was conducted in May at the National Hospital for Neurology and Neurosurgery and involved real-time analysis of camera footage by AI to identify critical anatomy in brain to be avoided.
The team of surgeons remained in full control of the event while AI was used to access and analyse live camera feed during the operation.
And this is where the AI tool proved its mettle, having trained on hundreds of hours of similar surgical videos of pituitary tumour removals.
The doctors were using data provided by AI based on a camera planted at the base of the patient’s skull via the nose.
Surgeons in London claimed success in performing the world’s first AI-assisted surgery to remove a tumour in the brain that saved the vision of a 48-year-old man. The operation was conducted in May at the National Hospital for Neurology and Neurosurgery and involved real-time analysis of camera footage by AI to identify critical anatomy in brain to be avoided.
A report published by The Guardian said the surgery was done two months ago but the details weren’t disclosed till last week when the patient was declared to be fully recovered. The team of surgeons remained in full control of the event while AI was used to access and analyse live camera feed during the operation.
“It was involved in helping identify key structures such as nerves and blood vessels that should be avoided near the 11mm (0.4in) tumour on the patient’s pituitary gland,” the paper said, while describing that this delicate part of the brain is located next to the arteries that supply it with blood as well as the optic nerves that control vision.
Going a millimetre either way could have made a critical difference, health officials told the publication and revealed that any error could have resulted in one of blindness, stroke and death. In fact, unearthing the tumour at a location that was hidden from view also posed a greater challenge as surgeons would’ve had to cut a precise path to get to the benign growth.
And this is where the AI tool proved its mettle, having trained on hundreds of hours of similar surgical videos of pituitary tumour removals. The AI system was exposed to a large sample size of surgical examples that a surgeon would have taken several years to come across, according to Sophia Bano, associate professor in Robotics and AI at the University College.
During its training, researchers manually drew around the vessels and nerves in each video so the AI could get an accurate idea of where they are, Bano told The Guardian. The doctors were using data provided by AI based on a camera planted at the base of the patient’s skull via the nose. IT was tracking the surgical instruments of the doctor in real time.
The hospital also noted that it had previously used the technology as a research tool but the operation was the first time it had been used on a patient. “It is designed to help recognise critical anatomy, surgical instruments and tissue interactions in real time, supporting the surgeon during highly delicate procedures,” Bano was quoted as saying.
