Researchers at Johns Hopkins reported an AI-driven method that measures adrenal gland size on routine chest CTs to quantify chronic stress, introducing what they say is the first imaging-based biomarker for long-term stress exposure. Presented at the RSNA meeting, the deep-learning model automatically segments the adrenal glands and produces an Adrenal Volume Index, which correlated with self-reported stress, cortisol levels and future cardiovascular outcomes in nearly 3,000 participants from the MESA cohort. The team says the approach could mine millions of existing scans for latent risk signals, but cautions the findings need external validation across populations, scanners and ages before clinical use. The work underscores how AI could turn standard imaging into a low-friction screening tool for stress-related disease, a growing public-health concern.
Related articles:
— Stress and Heart Health
— Adrenal gland
— Cortisol
— Artificial intelligence in healthcare
