Siberian Researchers Develop World's First AI Model for Diagnosing Depression
The system achieves up to 93% accuracy and could become a valuable decision-support tool for psychiatrists.
Researchers from Tomsk State University and Novosibirsk State University have developed what they describe as the world's first AI model for diagnosing depression. According to the team, the system achieves up to 93% accuracy. The researchers say they are the first to combine electroencephalography (EEG) data with genetic markers in a single algorithm. These two data sources are typically analyzed separately, and the team says their combination is what gives the model its high accuracy.
"The algorithm's main advantage is that it can analyze a new patient and determine the probability of depression with a high degree of accuracy. The computer can identify complex combinations of indicators that a psychiatrist simply cannot detect on their own. The important point is that we are not replacing psychiatrists - we are giving them a tool that helps eliminate subjectivity," said Neda Firoz, junior researcher at the High Energy Physics Data Analysis Laboratory, Faculty of Physics, Tomsk State University.
The AI model was trained using data from nearly 400 people.
