GENERICO.ruScienceScientists have learned to recognize depression using artificial intelligence

Scientists have learned to recognize depression using artificial intelligence

A computer neural network was trained to distinguish truly ill patients from healthy ones using MRI images

Scientists from the Baltic Federal University named after M.V. I. Kant on certain indicators of the nervous system of patients.

Computer neural network trained to distinguish really sick from healthy people by MRI images

Big depressive The disorder, which is associated with loss of interest in activities, insomnia, drowsiness and guilt, affects about 280 million people worldwide.

However, it is often difficult to make an accurate diagnosis. After all, the doctor most often relies on a subjective assessment of his condition by the patient. They decided to call on machine learning algorithms, that is, artificial intelligence, to help, but it had to be first trained on a large number of truly sick and healthy people.

For this, the authors took snapshots of activity using functional magnetic resonance imaging the brains of already identified patients with depression and healthy people, and then built functional networks on their basis. After that, according to the picture reflecting the interaction of different parts of the brain, the intensity of blood flow in them, a classification of sick and healthy people was built.

Based on the networks obtained, scientists compared brain function in 35 patients with depressive disorder and in 50 healthy people, and then, using machine learning methods, the signs of both groups were separated.

The proposed approach made it possible to distinguish sick and healthy people with an accuracy of 82.6%.


Please enter your comment!
Please enter your name here

LATEST POSTS in this category