Could AI boost brain surgery outcomes?
Health & Wellbeing
The AI diagnostics system can detect residual cancerous mass in seconds
Spotted: In 2019, 347,992 new cases of brain cancer were recorded around the world. Because it’s so hard to treat, the disease has a low survival rate, as well as a high economic burden. Neurosurgeons are rarely able to remove the whole mass of a life-threatening brain tumour because they aren’t always able to differentiate healthy tissue from the cancer. Now, a new AI diagnostics system may be able to help with this challenge.
The diagnostics system, called FastGlioma, has been developed by researchers from the University of Michigan and University of California San Francisco. The team claims that in tests, the AI-powered model outperformed existing methods for identifying tumour remains by a wide margin, and was able to detect residual mass in just 10 seconds. At the moment, there are limiting factors to current methods of detection. For instance, MRI imaging can only be used if a machine is available, and a fluorescent imaging agent can be used during surgery, but this method doesn’t work for all tumour types.
FastGlioma works by combining microscopic optical imaging with an AI foundation model that was pre-trained using over 11,000 surgical specimens and four million unique microscopic fields of view. High-resolution imagery took 100 seconds to acquire and the AI was able to correctly classify residual tumour and healthy tissue 92 per cent of the time. Lower-resolution images can be acquired in just 10 seconds, but the accuracy rate drops to 90 per cent.
Todd Hollon, M.D., a neurosurgeon at University of Michigan Health and assistant professor of neurosurgery at U-M Medical School states that: “The technology works faster and more accurately than current standard of care methods for tumour detection and could be generalised to other paediatric and adult brain tumour diagnosis. It could serve as a foundational model for guiding brain tumour surgery.”
The FastGlioma technology will be applied to other cancers such as prostate, breast, lung, head, and neck cancers in future studies.
Written By: Jessica Wallis
28th November 2024
Email: MichMedMedia@med.umich.edu
Website: michiganmedicine.org
Contact: michiganmedicine.org/contact