Artificial Intelligence Devices Help Doctors Detection Of Aggressive Brain Tumors Quickly

Kun-Hsing Yu, a senior researcher in brain cancer. (photo; twitter @HarvardDBMI)

JAKARTA - An artificial intelligence tool (AI) has emerged to assist doctors in fighting aggressive brain tumors. This tool helps identify important characteristics that provide guidance during surgery.

Cryptosection Histopathology Assessment and Review Machine (Charm) is an advanced tool that efficiently analyzes images to identify the genetic profile of glyomas, an aggressive type of brain tumor. Currently, this process takes days or weeks. Kun-Hsing Yu, senior author of the report on July 7 on Med, explains that surgeons rely on detailed diagnosis to guide their operations.

Although the accuracy of this tool is not yet comparable to current genetic tests, it can quickly predict tumor profiles. According to Yu, this rapid analysis allows doctors to resume proper treatment without the need to schedule and perform additional operations, thus saving valuable time.

In addition, Charm can distinguish between malignant and tame tumor cells and determine the level of tumor ferocity. This is an assessment usually carried out by human pathologists during surgery. However, according to Yu, Charm can eliminate the need to wait 10 to 15 minutes or the presence of a pathologist who is ready during surgery.

Glioma, especially the aggressive strain known as glioblastoma, is a significant threat, with untreated cases causing death in less than six months. According to the American Faculty Surgery Association, unfortunately, only 17% of individuals were diagnosed with glioblastoma who survived beyond the second two years.

Yu and his team trained machine learning algorithms using images of brain surgery samples and validated their accuracy with patient diagnosis. Charm is better at identifying tumor genetic profiles than other AI systems.

When making important decisions about the extent to which tissue is removed and the use of drug-coated suppliers in treating glyoma tumors, surgeons rely heavily on tumor genetic profiles. Unfortunately, this information is currently taking a long time.

Research conducted by Yu and his team is part of various initiatives that use AI to improve diagnosis and treatment of cancer. In an editorial published in the June issue, Lancet Oncology, the ability of certain systems is emphasized because of its ability to accurately identify individuals with high risk of developing pancreatic, lung, and breast cancer.