Mapping the Genetic Blueprint of the Corpus Callosum: AI and Imaging Insights

Researchers at USC’s Mark and Mary Stevens Neuroimaging and Informatics Institute have used AI-driven imaging genetics to map the corpus callosum’s genetic architecture, identifying dozens of genomic regions that shape callosal size and thickness and highlighting prenatal neurodevelopmental pathways relevant to neurodevelopmental and psychiatric disorders.

Earlier imaging-genetics efforts relied on manual, heterogeneous segmentation and smaller, clinic-based cohorts that limited throughput and reproducibility. Here, an AI-enabled analysis applied to large MRI and genotype datasets standardized measurements across scan types and sites, increasing scale, consistency, and statistical power.

The team automated segmentation of callosal subregions, harmonized MRI measures, and implemented a genome-wide association framework using imaging and genotype data from more than 50,000 participants spanning childhood to late adulthood.

The core genetic result identified dozens of loci influencing corpus callosum morphology; several implicated genes show peak expression prenatally, implicating early neurodevelopmental processes and sharpening the temporal window for follow-up studies of corpus callosum genetics.