Peter Salovey President | Yale University
Peter Salovey President | Yale University
Yale researchers have utilized machine learning to model the interplay of neurobiological and environmental factors in shaping mental health in adolescents. This new study, published in the journal Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, demonstrates significant improvements in detecting and predicting mental health issues compared to existing methods.
The research was led by Erica L. Busch and May I. Conley, Ph.D. candidates in the Department of Psychology, under the supervision of Arielle Baskin-Sommers, an associate professor of psychology and psychiatry at Yale.
"Nearly 75% of all mental health disorders appear during adolescence, with about half occurring by age 14," noted the researchers. Developmental scientists have long sought to understand how neurobiological and environmental factors contribute to emotional and behavioral problems but often considered these factors in isolation or as simple interactions.
The study employed multi-view manifold learning, specifically developing an exogenous PHATE (E-PHATE) algorithm to model brain-environment interactions using data from the National Institutes of Health-supported Adolescent Brain and Cognitive Development (ABCD) Study. This technique allowed for bridging and visualizing brain images collected via fMRI with information on adolescents' environments, enabling predictions of individual differences in cognition and emotional and behavioral symptoms both at a snapshot in time and two years later.
"The brain–environment manifolds of certain brain regions — for example, frontoparietal and attention networks — vastly improved detection and prediction of mental health issues relative to prior state-of-the-art approaches," said Busch.
These findings highlight the potential of manifold learning techniques to enhance research on the neurobiology of emotional and behavioral problems in adolescents. "For a long time, developmental scientists have faced the challenge of testing theories that are hiding in plain sight," said Conley. "We recognize youth’s experiences in their environments and neurobiology both influence emotional and behavioral development."
The researchers also emphasized the importance of combining multiple variables characterizing adolescents’ environments into E-PHATE. They found greater correlations between brain activity and mental health symptoms when modeling either neighborhood or familial environments but saw further improvements when combining these metrics with others.
"It is important that we improve our ability to capture the complex transactions between the person and their environment," said Baskin-Sommers. "New methods are needed to handle multiple types of data and estimate their interactions within individuals."
The interdisciplinary collaboration behind this method exemplifies how complex transactions can be estimated effectively.