Using information from routinely collected electronic health data, Duke researchers discovered that knowing whether a child has an early history of language or developmental delay improved their ability to predict whether that child will later be diagnosed with autism.
This work was conducted by AI Health fellow Angel Huang, PhD, and supervised by Benjamin Goldstein PhD and Matthew Engelhard, MD PhD, and published in the Journal of Biomedical Informatics. This work was supported by grants from the National Institutes of Health (NIH) Autism Centers of Excellence and National Institute of Mental Health.
Wei A. Huang, Matthew Engelhard, Marika Coffman, Elliot D. Hill, Qin Weng, Abby Scheer, Gary Maslow, Ricardo Henao, Geraldine Dawson, Benjamin A. Goldstein, A conditional multi-label model to improve prediction of a rare outcome: An illustration predicting autism diagnosis, Journal of Biomedical Informatics, Volume 157, 2024, 104711, ISSN 1532-0464, https://doi.org/10.1016/j.jbi.2024.104711.