Fall 2024 Director Updates

As I was writing the column, I was reminded that World Mental Health Day is October 10, 2024. This year's theme is “Mental Health at Work.”  In addition, there is a great deal of current discussion about AI's potential strengths, not just its pitfalls. The objective of this column is thus to show the potential impact of AI on mental health interventions. 

The Minnesota Robotics Institute has many faculty working in this exciting area. For example, Professors Gini, Zhao, Esler, Bernstein, Lim, Cullen, Conelea, Su, Cullen, and Morellas have produced remarkable results using robotics for interventions in conditions such as autism and Tourette’s. A group of us, including Drs. Esler,  Simacek, Elison, Morellas, Bedros, and Morris, are looking at how machine learning can be used to measure the progress of neurodevelopmental disorders (autism, ADHD, etc.). The faculty involved spans three UMN colleges: the College of Science and Engineering, the Medical School, and the College of Education and Human Development. Our interdisciplinary approach, based on generative AI, has the potential to revolutionize mental health care by enabling timely and adaptable interventions, which are crucial in the treatment of these conditions.

By facilitating accurate measurements of the progress of these disorders, this technology could significantly reduce their personal and societal burdens. Many individuals with these disorders currently wait months for a diagnosis or for an evaluation of an intervention’s effectiveness, which often leads to more severe manifestations over time. Early identification and intervention could pave the way for more targeted cognitive behavioral therapies, potentially averting such escalations. This tool could also help mitigate the need for pharmaceutical interventions at advanced stages of these disorders, reducing the risk of severe side effects associated with long-term medication use. The mental health literature consistently emphasizes the positive impact of early intervention, and our work aligns perfectly with that paradigm.

A key advantage of this research is its potential to generate patient-specific data, enabling the creation of tailored treatment plans. This personalized approach could offer precision phenotyping and a method for measuring response to treatment— in contrast to current methods, where treatment decisions are often based on limited empirical data and individual clinician experience. By providing a more data-driven framework, this tool could become an essential component in developing both effective and cost-efficient mental health strategies.

While initially focused on autism, the applications of this work could extend to other neurodevelopmental disorders for which there is a benefit in measuring the disorder’s progress. The project presents unique challenges in real-time data processing and analysis, making it an ideal testbed for innovations in embedded systems. Particularly intriguing is the need to detect subtle visual behaviors associated with these disorders, pushing the boundaries of current computer vision and machine learning capabilities beyond traditional applications of human monitoring. 

Have a great academic year,

Nikos Papanikolopoulos
Minnesota Robotics Institute Director