New AI approaches are connecting development, environment, and disease
Researchers are using artificial intelligence to uncover links between early human development, environmental exposures, and later-life diseases. These new AI approaches analyze large datasets to identify patterns that traditional methods might miss, potentially leading to earlier interventions and better understanding of disease origins.
Background
- This piece covers how researchers are using AI to study the "exposome" — the totality of environmental exposures (pollution, diet, stress, chemicals) a person experiences from conception onward, and how that interacts with development to shape disease risk.
- Traditional epidemiology struggles to link complex, mixed exposures to health outcomes. New AI/ML techniques can find patterns in large, messy datasets (e.g., combining health records, wearable sensor data, satellite pollution maps, and genomic data).
- Key approaches include: deep learning to model non-linear interactions between multiple exposures; natural language processing to extract exposure data from clinical notes; and causal inference methods to distinguish correlation from causation in observational data.
- Why it matters: Most chronic diseases (cancer, diabetes, asthma) arise from gene-environment interactions we poorly understand. Better AI-driven exposome analysis could enable earlier intervention, personalized prevention, and more targeted public health policy.