AI Deciphers Tissue Images to Predict Age and Reveal Genetic Links

Histological images, detailed snapshots of tissue structure, are rich sources of information, yet fully extracting their potential, especially outside cancer research, remains a significant challenge. A new study addresses this by introducing a powerful statistical framework designed to link these complex visual data with a sample's underlying genetic makeup, gene activity, and chronological age.

The research team first established a connection between image features and genotypes, pinpointing 906 "image quantitative trait loci" (imageQTLs) — specific genetic regions significantly associated with image characteristics. They further identified differentially expressed genes by grouping samples based on similar image features.

Building on this, a deep-learning model was developed to accurately predict gene expression in specific tissues directly from raw images and their features, highlighting gene sets associated with observed morphological changes. The study also constructed a deep-learning model to predict chronological age. As lead author Meng notes in the paper, "Finally, we constructed another deep-learning model to predict chronological age directly from raw images and their features, revealing relationships between age and tissue morphology, especially aspects derived from nucleus features." This ability to discern age-related changes from tissue images opens up exciting possibilities for understanding the biological mechanisms of aging and its impact on tissue health.

Both deep-learning models are supported by an innovative computational approach that effectively compresses massive gigapixel whole-slide images while extracting detailed, interpretable nucleus features, integrating both large-scale tissue morphology and smaller local structures. By making all interpretable nucleus features, identified imageQTLs, differentially expressed genes, and the deep-learning models publicly available, the researchers aim to accelerate further investigations. This comprehensive approach, linking microscopic tissue details to genetic, transcriptional, and age-related insights, marks a significant step forward in leveraging the full power of histological data for advanced biological and medical research.