Revolutionary AI Tool Enables Real-Time Molecular Diagnosis of Brain Tumors During Surgery

Scientists have unveiled a groundbreaking artificial intelligence (AI)-based tool that has the potential to revolutionize brain tumor surgeries. The tool, known as CHARM (Cryosection Histopathology Assessment and Review Machine), can rapidly decode a brain tumor’s DNA, providing neurosurgeons with crucial molecular information during surgery.

Thе ability to dеtеrminе thе molеcular idеntity of a tumor in rеal timе allows surgеons to makе critical dеcisions on thе spot. With this information, they can accurately gauge the extent of brain tissue to be removed and even consider administering tumor-killing drugs directly into the brain during the operation. By making these decisions during surgery, patient outcomes can be significantly improved.

Currently, the process of profiling a tumor’s molecular type can take several days to weeks, causing delays in treatment planning. This time-consuming approach also carries risks. Removing excessive brain tissue when dealing with a less aggressive tumor can lead to neurologic and cognitive impairment. Conversely, if too little tissue is removed for a highly aggressive tumor, malignant cells may be left behind, leading to rapid growth and spread.

However, a team of researchers led by Kun-Hsing Yu, Assistant Professor of biomedical informatics at the Blavatnik Institute at Harvard Medical School, has designed CHARM to overcome these challenges. The tool extracts previously untapped biomedical signals from frozen pathology slides, enabling molecular profiling of tumors during surgery.

“Our innovative solution tackles this obstacle head-on by harnessing previously unexplored biomedical signals from cryogenic pathology slides,” elaborated Professor Yu. “By enabling the extraction of these signals, our tool empowers surgeons to achieve intraoperative molecular diagnosis in real-time, revolutionizing the advancement of precision oncology during surgical procedures.”

CHARM has shown impressive accuracy in its initial tests. This groundbreaking advancement was meticulously crafted through the analysis of an extensive dataset consisting of 2,334 brain tumor samples obtained from a diverse pool of 1,524 individuals diagnosed with glioma. By encompassing a wide range of patient populations, our development process ensured robustness and reliability in our findings, setting a new standard in glioma research and diagnosis. When tested on previously unseen brain samples, CHARM achieved a remarkable 93% accuracy in distinguishing tumors with specific molecular mutations. Moreover, it successfully classified the three major types of gliomas, each with distinct molecular features, providing valuable prognostic information and guiding treatment decisions.

The tool also excels in capturing visual characteristics of the tissue surrounding malignant cells. It can identify areas with higher cellular density and increased cell death, indicating more aggressive types of gliomas. Additionally, CHARM can pinpoint clinically significant molecular alterations in low-grade gliomas, a less aggressive subtype of glioma with differing growth patterns and responses to treatment.

Although CHARM’s development and testing focused on glioma samples, the researchers believe it has the potential to be retrained to identify other brain cancer subtypes as well. Howеvеr,  bеforе its dеploymеnt in hospitals,  thе tool must undеrgo rigorous clinical validation and rеcеivе clеarancе from thе US Food and Drug Administration (FDA). 

The findings of this study, published in the esteemed journal Med, signify a significant step forward in brain tumor surgery. If CHARM proves successful in real-world settings and gains regulatory approval, it could potentially transform the field of neurosurgery and advance the era of real-time precision oncology.

Leave a comment