Infant neuroimaging and prediction: towards early identification and individualized intervention

• In Event: Using Connectomics and Machine Learning to Examine Brain Development

Abstract

The first two years of life mark a rapid, dynamic period of early postnatal brain development that coincides with the emergence of burgeoning cognitive abilities. It is during this same period that atypicalities in development associated with developmental delay and neurodevelopmental disorders, including autism spectrum disorder (ASD), are first apparent. Initiating presymptomatic intervention during a period of heightened neuroplasticity is hypothesized to have the greatest long-term benefits for the child. Such an approach relies heavily on the ability to accurately detect at-risk infants and identify personalized targets for intervention.

Here we will present a series of published (Hazlett et al., 2017; Emerson et al., 2017; Girault et al., 2019) and preliminary data that seeks to move the field towards early identification and individualized intervention in the first two years of life for at-risk children. We couple pediatric neuroimaging with machine learning approaches to predict categorical (i.e. diagnostic) and continuous (i.e. cognitive) outcomes at 24 months of age using data derived from magnetic resonance imaging (MRI) scans taken in the first year of life. Study populations include typically-developing infants, infants born pre-term, and infants at high familial risk for ASD.

We will highlight best practices and approaches in the emerging field integrating machine learning and MRI for prediction (Girault & Piven, 2020; Mostapha & Styner 2019), define the utility of a clinically actionable diagnostic prediction of ASD, and convey the role that continuous outcome prediction will play in the development of targeted interventions. We also discuss the ethical, legal, and social implications of the presymptomatic prediction of later outcomes during infancy.

Authors

• Jessica Girault, University of North Carolina at Chapel Hill
  Presenting Author

• Mark D. Shen, University of North Carolina at Chapel Hill
  Non-Presenting Author

• Martin Styner, University of North Carolina at Chapel Hill
  Non-Presenting Author

• Joseph Piven, University of North Carolina at Chapel Hill
  Non-Presenting Author

• John R. Pruett, Washington University in St. Louis
  Non-Presenting Author