UIC researcher to test voice-activated AI to manage mental health symptoms

Researchers at University of Illinois Chicago are studying a novel approach to delivering care to those with moderate depression and anxiety: through artificial intelligence, or AI.

Dr. Jun Ma, professor of academic internal medicine and geriatrics associate head of scholarly activity, department of medicine, Beth and George Vitoux Professor of Medicine and director of Vitoux Program on Aging and Prevention.

Dr. Jun Ma, Beth and George Vitoux Professor of Medicine at the UIC department of medicine, has received a $2 million grant from the National Institutes of Mental Health for the “Study of a PST-Trained Voice-Enabled Artificial Intelligence Counselor (SPEAC) for Adults with Emotional Distress.”

The first part of the two-phase, five-year project will develop and test a voice-enabled, AI virtual agent named Lumen, trained to deliver Problem Solving Therapy (PST), for patients with moderate, untreated depressive and/or anxiety symptoms. This first phase is awarded for two years.

“The central motivation is trying to leverage the latest technology as a vehicle to improve access to proven psychotherapy,” Ma said.

Using the same technology as Amazon’s Alexa, Ma and co-principal investigator Dr. Olusola Ajilore, UIC associate professor of psychiatry, plan to develop an app and install it on iPads that study participants will be given to use. The Lumen app will act as a virtual mental health agent, talking through problem-solving steps and strategies with the participants following a validated treatment protocol.

Problem-solving therapy is a well-proven, structured approach to help people focus on learning cognitive and behavioral skills, Ma said.

Individuals will complete eight, one-on-one counseling sessions over 12 weeks. During each session, they identify a problem they view as affecting their life and a source of emotional distress, and the counselor coaches them to define goals and possible solutions. Solutions are compared and an action plan is made to implement the solution chosen. Lumen will be programmed using the Alexa Skills Kit to act as the virtual counselor working with study participants, taking them through problem-solving steps and counseling them to think about ways to engage in meaningful and enjoyable activities in order to improve their emotional well-being.

During the first phase, 80 study participants who report elevated depressive and anxiety symptoms will test the newly developed Lumen. If the results are promising, then the project will progress to the second phase for another three years for further testing with 200 additional participants.

“This project’s take-away message is not whether this agent or any other AI digital health agent would replace mental health specialists, but may improve the reach and impact of psychotherapy — complement and extend existing mental health services which, as we know, are severely in short supply relative to the demand,” Ma said.

“Our ultimate goal is to take advantage of accessible technology to lower the barriers to mental health care for those who need it the most,” Ajilore said.

The study’s first phase also will investigate Lumen’s ability to engage target neural mechanisms of action — amygdala for emotional reactivity and dorsal lateral prefrontal cortex (DLPFC) for cognitive control. Study participants will undergo functional MRI tests prior to engaging with Lumen, and 14 weeks after.

There is much emerging research on using AI to promote health and wellness, Ma said. However, this project will be the first to combine AI voice technology and the well-proven problem-solving treatment and measure the neural mechanisms in relation to symptom improvement in depression and anxiety.

The study is supported by grant from the National Institutes of Mental Health (1R61MH119237-01A1).

Co-investigators are Philip Yu, professor of computer science, and Dr. Ben Gerber, professor of academic internal medicine and geriatrics, both at UIC; Dr. Joshua Smyth, professor of biobehavioral health and medicine at the Pennsylvania State University; Dr. Thomas Kannampallil, assistant professor of anesthesiology at Washington University in St. Louis, and Lan Xiao, biostatistician at Stanford University.