Artificial intelligence tool decodes the brain cancer genome during surgery to remove it.

Real-time tumor profiling can guide surgical treatment decisions. The research was published in the journal Med.

The scientific team, led by Harvard Medical Schooldesigned an artificial intelligence tool that can rapidly decode the DNA of a brain tumor during surgery to determine its molecular identity.

This information is critical, as with the current approach it can take anywhere from a few days to a few weeks to collect.

Knowing the molecular formula of a tumor allows neurosurgeons to make decisions such as how much brain tissue to remove and whether to give anticancer drugs directly into the brain at the time of surgery.

Beyond decisions made during surgery, knowing the molecular makeup of a tumor provides clues about its aggressiveness, behavior, and potential response to various treatments.

The tool is called CHARM (Cryosection Histopathology Assessment and Review Machine) and developed using 2,334 brain tumor samples from 1,524 people with glioma. When tested, it discriminated tumors with specific molecular mutations accurately 93% and successfully classified three main types of gliomas with distinct molecular features.

He also successfully recorded the optical characteristics of the tissue surrounding the malignant cells and identified clinically relevant molecular changes in a glioma subtype that is less aggressive and therefore less likely to invade the surrounding tissue.

Finally, he linked the appearance of the cells with the molecular profile of the tumor and this means that the algorithm can pinpoint how a cell’s appearance relates to a tumor’s molecular type.

The tool is freely available to other researchers. However, its value still needs to be confirmed clinically through real-world trials and approved by the FDA before it can be used in hospitals, the research team notes.

Furthermore, the researchers point out that while the model was trained and tested on glioma samples, it could be successfully retrained to identify other brain cancer subtypes as well.

“The ability to determine intraoperative molecular diagnosis in real time during surgery may advance the development of real-time precision oncology,” notes lead study author and assistant professor of Biomedical Informatics at Harvard Medical School Kun-Hsing Yu.