Artificial Intelligence From IBM Can Predict Schizophrenia Simply By Studying The Blood Flow In The Brain - Alternative View

Artificial Intelligence From IBM Can Predict Schizophrenia Simply By Studying The Blood Flow In The Brain - Alternative View
Artificial Intelligence From IBM Can Predict Schizophrenia Simply By Studying The Blood Flow In The Brain - Alternative View

Video: Artificial Intelligence From IBM Can Predict Schizophrenia Simply By Studying The Blood Flow In The Brain - Alternative View

Video: Artificial Intelligence From IBM Can Predict Schizophrenia Simply By Studying The Blood Flow In The Brain - Alternative View
Video: Artificial Intelligence Unleashed on the Brain: Using Machine Learning to Predict Alzheimer's 2024, March
Anonim

Schizophrenia is not a particularly common mental disorder in America, affecting only 1.2 percent of the population, or about 3.2 million people, but the consequences can be dire. However, groundbreaking research by IBM and the University of Alberta will soon help doctors diagnose the onset and possible severity of symptoms using simple MRI scans and a neural network designed specifically to study blood flow in the brain.

“This unique, innovative, multidisciplinary approach opens up new opportunities and contributes to our understanding of the neurobiology of schizophrenia, which can help improve the treatment of this disease,” said Dr. Serdar Dursun, professor of psychiatry and neurology at the University of Alberta.

The research team first trained their neural network on a database of 95 members, which included scans of both schizophrenic patients and a healthy control group. These images illustrated the flow of blood through different parts of the brain as patients performed a simple audio-based exercise. From this data, the neural network created a predictive model of the likelihood that the patient suffered from schizophrenia, based on the study of the characteristics of blood flow. The AI was able to accurately distinguish between the control group and patients with schizophrenia in 74% of cases.

"We have found a number of serious abnormal brain connections that can be investigated in the future," Dursun continued, "and the models created by AI are bringing us closer to finding objective indicators based on neuroimaging that are diagnostic and prognostic markers of schizophrenia."

Moreover, the model was also able to predict the severity of symptoms for the identified disease. These ideas could lead researchers to more effective diagnostic tools and treatment options. And why not? IBM's most famous AI, Watson, has already proven that neural networks can come up with effective cancer treatments.

Serg kite