New artificial intelligence tool could help doctors treat rare cancers more effectively
A new artificial intelligence (AI) tool could help doctors treat rare cancers more effectively by grading their severity with high precision.
Developed by researchers at the Royal Marsden Hospital and Institute of Cancer Research in London, the tool is almost twice as efficient at grading the aggressiveness of retroperitoneal sarcoma, a rare cancer that occurs in the connective tissue at the back of the abdomen. A study suggests the same approach as the current one.
By recognizing details not visible to the naked eye, the AI achieved an accuracy rate of 82 percent, compared to 44 percent for laboratory analysis.
The researchers say this tool could improve treatment and benefit thousands of patients with rare cancers each year. They are also excited about the tool’s potential for early detection of other cancers.
The AI tool works by using a technique called radiomics to identify signs of retroperitoneal sarcoma that are not visible to the naked eye on patient scans. This data is then used to train AI algorithms to grade the aggressiveness of the cancer.
The researchers tested the AI tool on 89 patients with retroperitoneal sarcomas in European and U.S. hospitals. They found that the tool was able to grade the aggressiveness of the cancer more accurately than a biopsy, which involves analyzing a small portion of cancerous tissue under a microscope.
The researchers believe that this AI tool could help doctors develop more personalized treatment plans for patients with retroperitoneal sarcoma. High-risk patients could receive specific treatments, while low-risk patients could avoid unnecessary treatments and follow-up scans.
The researchers now plan to further trial this AI tool on more patients with retroperitoneal sarcoma and other types of rare cancers. They hope that the tool will eventually be used around the world to help doctors improve the treatment of patients with rare cancers.
