Discovering the gene-brain-behavior link in autism via generative machine learning
This preclinical study introduces a novel AI-powered technique (3D TBM) that can accurately identify specific structural brain changes linked to a genetic variation (16p11.2 CNV) often found in autism, with 89-95% accuracy from brain images alone. These identified brain patterns are associated with articulation disorders and explain a portion of IQ variability, though the study cannot establish direct causality. The authors acknowledge that these findings require further clinical validation.