NHS trusts in England will be offered potentially game-changing artificial intelligence to help speed up cancer treatment.
OSAIRIS was developed by researchers at Addenbrooke’s Hospital in Cambridge in collaboration with Microsoft. It examines cancer patients’ scans to calculate where radiotherapy beams should be directed and ensure healthy organs are protected.
For humans, it’s a skilled and painstaking process, taking trained doctors between 20 minutes and three hours to perform for each patient. The same specialists working with artificial intelligence can do it about two and a half times faster, the researchers say.
With the AI boasting a 90% accuracy rate, the doctors give a second opinion, rather than undertaking the main bulk of the work themselves, freeing up their valuable time and reducing waiting times for patients. Researchers say that the scans go unamended around two-thirds of the time.
“OSAIRIS does much of the work in the background so that when the oncologist sits down to start planning treatment, most of the heavy lifting is done,” says Dr. Raj Jena, an oncologist at the Cambridge Universities Hospital NHS Foundation Trust, who led the research.
For cancer patients, the importance of this accuracy can’t be overstated, as protecting healthy organs from radiation is imperative to quality of life, post-treatment. If treatment is targeting the prostate, for example, avoiding damaging the nearby bladder and rectum is important to avoid continence issues going forward.
“That can get so bad that a patient’s life becomes dominated by that,” Dr. Jena told BBC Newsnight. “I know patients where they’ve got a map of the cities that they’re going to, so they know where all the loos are.”
Medical advancement is one of the most exciting fields in artificial intelligence, and while this particular use is for speeding up work and freeing up professionals’ time, other implementations rely on AI’s ability to spot trends that may go unnoticed by human eyes.
An AI trained to tell the difference between cardiomyopathy and constrictive pericarditis, for example, was able to achieve a success rate of up to 96% in two months, while trained doctors were hovering around 56%.
