Age-Stratified Individual-Specific Brain Features Improve Prediction of Autism Spectrum Disorder Behavioral Symptoms

This neuroimaging study analyzed fMRI data from 437 participants using a computational algorithm (COBE) to extract age-specific brain connectivity patterns in individuals with autism and controls. Age-stratified individual-specific features significantly improved prediction accuracy of behavioral symptoms, particularly for social behavior in adolescents with ASD, showing 41% average improvement over standard connectivity measures.