Home » Artificial Intelligence decodes tumors in the operating room

Artificial Intelligence decodes tumors in the operating room

by admin
Artificial Intelligence decodes tumors in the operating room

MeteoWeb

Artificial Intelligence is now able to decode the DNA of brain tumors in real time, directly in the operating room: in this way it can guide the work of surgeons by suggesting how much tissue to remove and which treatments to administer to patients, thus bringing personalized medicine closer tailored to the characteristics of each tumor. The innovative tool, called Charmwas developed by a group led by Harvard Medical Schoolwhich published the result achieved in the journal Med.

It has been freely made available to other researchers, but has yet to be tested in real situations and undergo the necessary approvals. Currently, the standard approach involves taking brain tissue during surgery and freezing it for examination under a microscope. The main disadvantages, in addition to the time required for this procedure, are that freezing the tissue tends to alter the appearance of the cells and that even the most powerful microscopes struggle to detect all the subtle variations that occur in the DNA of tumor cells.

The researchers coordinated by Kun-Hsing Yu they then used AI to overcome these challenges. The study authors trained Charm using 2,334 brain tumor samples taken from 1,524 people. The device, when put to the test with never-before-seen samples, was able to distinguish tumors with specific mutations with 93% accuracy, but that’s not all. He has also been able to identify the characteristics of the tissue surrounding the malignant cells, which can signal the presence of more or less aggressive tumours. Charm will need to be periodically retrained to keep up with new knowledge: “Just like doctors – commenta Yu – who must engage in continuous training”.

Follow MeteoWeb on Google News

You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More

Privacy & Cookies Policy