Imagine waiting nearly a year for a diagnosis that could change your child’s life. For many families in rural areas like Missouri, this is a harsh reality when it comes to autism evaluations. But what if artificial intelligence could slash that wait time dramatically? Researchers from the University of Missouri School of Medicine believe it can.
In a groundbreaking study, lead author Dr. Kristin Sohl and her team partnered with Cognoa, Inc. to test CanvasDx, an FDA-approved medical device designed to assist primary care clinicians in diagnosing autism. This innovative tool uses AI algorithms to analyze patient data and predict whether a child is likely to have autism. If the data isn’t conclusive, it provides an 'indeterminate' result, ensuring accuracy over haste. And here’s where it gets controversial: while AI can’t replace human expertise, it’s proving to be a game-changer in areas where autism specialists are scarce.
But here’s where it gets even more intriguing: CanvasDx isn’t just about speed. It’s about accessibility. Families in rural Missouri often face a 97-mile journey to reach specialty care centers, a burden that delays diagnosis and treatment. By keeping evaluations local and leveraging AI, families can save time, money, and stress, receiving diagnoses up to 7 months earlier. This isn’t just a technological advancement—it’s a lifeline for underserved communities.
In the study, CanvasDx produced determinate results for 52% of the 80 children evaluated, with zero false positives or negatives. It also never contradicted a clinician’s diagnosis, underscoring the importance of trained professionals in the process. And this is the part most people miss: AI isn’t here to replace clinicians; it’s here to support them, providing objective data that streamlines evaluations and ensures timely, high-quality care.
Dr. Sohl emphasizes, ‘Identifying autism early and starting individualized support are critical for optimizing outcomes. Autistic children and their families deserve timely access to local expertise, and AI-integrated tools like CanvasDx can make that a reality.’ But here’s a thought-provoking question: As AI becomes more integrated into healthcare, how do we balance innovation with the human touch that’s so essential in diagnosing and treating complex conditions like autism?
What do you think? Is AI the future of autism diagnosis, or does it risk dehumanizing a process that requires empathy and nuance? Share your thoughts in the comments—we’d love to hear your perspective!
Dr. Kristin Sohl is a pediatrician at MU Health Care and a professor of pediatrics at the Mizzou School of Medicine. She is also the founder and executive director of the ECHO Autism program and the Medical Director of the Missouri Telehealth Network and the Office of Continuing Education for Health Professionals.
Source: Sohl, K., et al. (2025). Integration of an Artificial Intelligence–Based Autism Diagnostic Device into the ECHO Autism Primary Care Workflow: Prospective Observational Study. JMIR Formative Research. doi:10.2196/80733. Read the study.*
