Artificial intelligence tools show potential to help doctors fight aggressive brain tumors by identifying features that guide surgery.
The tool, called the Frozen Section Histopathology Evaluation and Review Machine (CHARM), studies images to rapidly identify the genetic profile of a type of tumor called a glioma. It currently takes days to weeks, said Kun-Hsing Yu, lead author of the paper. The report, published Friday in the journal Med. Surgeons use detailed diagnostics to guide surgery, and being able to get a diagnosis faster could improve patient outcomes and avoid multiple surgeries, Yu said. rice field.
Gliomas vary in severity, but a progressive form called glioblastoma can be fatal within six months if left untreated. According to the American Association of Neurosurgeons, only 17% of people with glioblastoma are alive two years after diagnosis.
Surgeons use information about the genetic profile of glioma tumors when deciding how much tissue to remove from a patient’s brain and whether to implant cancer drug-coated wafers. To do. However, obtaining that information currently requires time-consuming testing.
Yu and his team trained a machine-learning algorithm to do this by showing pictures of samples collected during brain surgery and checking the results against the patient’s diagnosis. CHARM has learned to perform as well or better than other AI systems in identifying the genetic profile of tumors.
Although this tool is not as accurate as current genetic tests, the computer system can predict a tumor’s profile almost instantly. The rapid analysis allows doctors to proceed with appropriate treatment without spending extra time scheduling and performing another operation, Yu said.
By using the results in conjunction with other information, “clinicians may be able to make the right decisions on the spot,” says the assistant professor of biomedical informatics at Harvard Medical School, who led the study. Mr Yu said.
CHARM can also differentiate malignant cells from benign cells and identify tumor grade, a measure of tumor grade. These are calls that a human pathologist can make during surgery, but the tool could eliminate waiting times of 10 to 15 minutes, or the need for pathologists to wait during surgery, Yu said. said.
In a news release, the researchers said the study showed promise, but CHARM still needs to be tested in real-world settings.
Yu’s team’s research is part of a broader effort to use AI to improve cancer diagnosis and treatment. His editorial in the June issue of The Lancet Oncology highlighted the ability of some systems to accurately identify people at high risk for pancreatic, lung and breast cancer.
