The Department of Computer Science is excited to announce Dr. Brent Munsell’s research on “Brain Characteristics in the First Year Predict Early Autism Diagnosis” has been selected for publication in the science journal, Nature. His research used neuroimaging data acquired at six months and twelve months to predict if a 24 month child will have autism.
This research was funded by an R01 NIH grant titled “A Longitudinal MRI Study of Infants at Risk for Autism” that examines brain overgrowth patterns of infants with autism and how this overgrowth information can be used to predict the onset of autism prior to 24 months of age. This research is a collaborative effort between eight universities in the United States and Canada, named the Infant Brain Imaging Study (IBIS) network. For more information about IBIS: http://www.ibisnetwork.org/. The IBIS network can be further divided into four groups: 1) Clinical site, 2) Data coordinating core, 3) Image processing core, and 4) Statistical analysis core.
Traditionally, a child is diagnosed with autism at 24 months of age using a standard diagnostic testing criteria (i.e. a questionnaire completed by the physician and the parent). Instead of waiting until the child reaches 24 months in age, the research proposed in the Nature publication developed a computational model to identify infants that are likely to have autism at twelve months. The model compares gray matter morphometric information derived from magnetic resonance imaging (MRI) data at six and twelve months of age and detects brain overgrowth onset from autism. Thus, early detection will result in early physical and occupational therapies that can dramatically improve the child’s quality of life.
The computational model was designed and developed by Dr. Brent Munsell in the Machine Learning and Medical Image Analysis Lab at the College of Charleston. Dr. Brent Munsell collaborates with Dr. Martin Styner at UNC Chapel Hill and Dr. Guido Gering at NYU, who lead the IBIS Network Image processing core. For more information about our research labs, visit the department research page.