Gondy Leroy, professor of Management Information Systems (MIS) in the Eller College of Management at the University of Arizona, is the recipient of a new $1.5 million grant from the National Institute of Mental Health to create health information technology (HIT) to support non-expert clinicians in identifying children at risk for autism spectrum disorder (ASD), which affects 1 in 54 children in the U.S.
“ASD is a condition for which early diagnosis is crucial, allowing for early treatment and the best long-term outcome, but is unfortunately often late or missed entirely, in large part due to lack of trained clinicians,” says Leroy. “To address this problem, our project will leverage machine learning algorithms to mark the electronic health records of children at high risk for autism.”
The project will also use free text in electronic health records to identify observable behavioral characteristics aligned with ASD as defined in the Diagnostic and Statistical Manual of Mental Disorders. The result of these two components will be health information technology that will support non-expert clinicians in their evaluation of children who may be at risk of ASD. The HIT will support early referrals leading to early diagnosis and therapy and will be especially useful in settings where domain expertise is missing.
Leroy and her co-investigators—at the University of Arizona, Sydney Rice M.D., professor of pediatrics and public health; Jennifer Andrews, assistant professor of pediatrics; Maureen Kelly-Galindo, pediatric genetics nurse and clinical supervisor in the genetics counseling program; and Winslow Burleson, professor of information science; and at Phoenix Children’s Hospital, Richard Frye M.D., a child neurologist—will leverage their existing work with the goal of achieving higher accuracy and efficiency with ASD diagnosis.
“Our solution addresses critical barriers in diagnosing ASD and will support clinicians with varying expertise,” says Leroy. “The final product will support human-decision making by processing electronic health records and suggesting whether a child is at high risk or not with supporting information.”
The researchers also intend to evaluate their work by comparing accuracy, confidence and efficiency with and without the health information technology of children at risk for ASD versus other neurodevelopmental issues.