An artificial-intelligence tool called the Cryosection Histopathology Assessment and Review Machine (CHARM) has demonstrated significant potential in aiding doctors in the fight against aggressive brain tumors. This tool analyzes images to rapidly identify the genetic profile of glioma tumors, a process that usually takes days or weeks. The findings of a report released on July 14 in the journal Med highlight the potential of CHARM to provide detailed diagnoses during surgery, potentially improving patient outcomes and reducing the need for multiple surgeries.
Glioma tumors, particularly the aggressive form known as glioblastoma, can be fatal within six months if left untreated. Unfortunately, only 17% of people diagnosed with glioblastoma survive beyond the second year. Surgeons currently rely on genetic information about the tumor to guide their surgical decisions, including how much tissue to remove and whether to use drug-coated wafers. However, obtaining this information is a time-consuming process.
To address this challenge, Kun-Hsing Yu and his team of researchers trained a machine-learning algorithm with images of samples acquired during brain surgery. The algorithm, implemented in the CHARM system, was then compared against the patients’ diagnoses. CHARM demonstrated the ability to match or outperform other AI systems in identifying the genetic profile of tumors.
Although CHARM is not as accurate as current genetic tests, it can rapidly predict a tumor’s profile. This swift analysis could enable doctors to proceed with appropriate treatment without the need for additional surgeries or scheduling delays. Yu, an assistant professor at Harvard Medical School, asserts that using CHARM’s results alongside other clinical information empowers clinicians to make informed decisions on the spot.
In addition to its profiling capabilities, CHARM can distinguish between malignant and benign tumor cells and determine a tumor’s grade, indicating its level of aggressiveness. While these assessments are typically made by human pathologists during surgery, CHARM could eliminate the need for a 10-to-15-minute wait or having a pathologist on standby.
It is worth noting that CHARM’s effectiveness should be further evaluated in real-world settings, as stated by the researchers in a news release. Yu’s team’s work contributes to the broader application of AI in cancer diagnosis and treatment. A recent editorial in the Lancet Oncology emphasized the accuracy of some AI systems in identifying individuals at high risk of pancreatic, lung, and breast cancer. These advancements highlight the potential of AI in improving cancer care.
Credit: The Star : Tech Feed
