Skip to content Skip to sidebar Skip to footer

Source: Pourquoi docteur? as of 18-06-2020

The machine is capable of determining the severity level of a glioma with a reliability of almost 98%.

Artificial intelligence is now used in almost all fields, including health. In a  study by the Institute for Integrated Cell-Material Sciences,Researchers in India and Japan show that it could become a screening tool in the management of  cancers. The tool they have developed is able to estimate the severity of a glioma, in order to administer the most appropriate treatment.

Knowing the grade of the tumor: a necessity to choose the right treatment

Glioma  is a brain tumor: it affects glial cells, which protect and nourish  neurons. Its severity is generally classified into different categories, called grade. The higher the tumor, the more aggressive the tumor, the more difficult the treatment. Knowing the precise level of severity of glioma is the only method that allows doctors to provide the most appropriate treatment. Today, the diagnosis is made from MRIscans, sometimes deciphered by artificial intelligence. They are able to detect details invisible to the naked eye on the shape or texture of the tumor.

A highly accurate tool

The scientific team at the Institute for Integrated Cell-Material Sciences has developed artificial intelligence techniques to obtain an even more accurate tool than those commonly used. The researchers used THE MRIs of more than 200 people to test it. Some of these tests resulted in the glioma-classifying machine  gliomes being trained, and the rest was used to verify that the model was working.

“Our method has surpassed the advanced techniques used to predict the level of gravity of gliomas,”notes Balasubramanian Raman, data scientist and co-author of this research. According to their findings, artificial intelligence has a accuracy of 97.54%: which means that it finds the level of severity of the tumor, without error, in most cases. After these encouraging discoveries, scientists would like to develop a semi- or fully automatic machine using this method to help medical teams make their  diagnoses.

Show CommentsClose Comments

Leave a comment

News ORS © 2020. All Rights Reserved.