
Kazim Topuz, an associate professor of business analytics and operations management in The University of Tulsa’s Collins College of Business, has completed a study of MDscan. The promising new tool uses artificial intelligence to scan for 10 different types of mental disorders simultaneously.
Topuz co-authored the paper that reported results of the MDscan study, which was published in the Journal of Management Information Systems last month.
Nearly 1 billion people around the world suffer from mental disorders, such as depression and anxiety, and as many as 94% of those go undiagnosed. Topuz and the other contributors – from Washington University in St. Louis, Rochester Institute of Technology, University of South Florida, and the global Bosch Center for Artificial Intelligence – were examining MDscan’s efficacy as an initial assessment tool to address these untreated disorders.
MDscan is powered by AI and uses data from a standardized questionnaire called the SCL-90-R. It doesn’t replace the work of psychiatrists and counselors but instead helps identify who needs support.
MDscan uses advanced techniques, including the team’s ShapRadiation algorithm, to take data from the questionnaire and creates a visual representation that is similar to a diagnostic image or brain scan. Clinicians can then actually see the patterns and severity of symptoms. This can make diagnoses much faster and clearer.
Generally speaking, Topuz’s work focuses on designing probabilistic graphical models, including Bayesian Belief Networks and Markov Networks, integrated with data mining techniques for data-driven decisions, and explainable AI and its applications in health care, accident severity, student retention, and mental health. His work has been published in the Journal of Management Information Systems, European Journal of Operational Research, Decision Support Systems, Annals of Operations Research, and Journal of Business Research, among others.