Scientists in Austria have created an artificial intelligence method to estimate the biological age of individual human organs using tissue samples. The approach, published today in Nature Medicine, could lead to blood tests that track aging and disease across the body without invasive biopsies.
AI decodes aging from tissue architecture
A research team led by André Rendeiro at the Research Center for Molecular Medicine of the Austrian Academy of Sciences trained AI models on over 25,000 high-resolution histological images from 40 different tissue types. These images came from the Genotype-Tissue Expression Project, a large-scale effort to map human gene activity.
Their work focused on how tissue structure changes over time rather than molecular shifts. Age emerged as the strongest factor shaping these structural patterns, enabling the AI to predict biological age with a mean error of 4.9 years—better than existing DNA-based methods.
“Our tissues carry a remarkably detailed record of the aging process,” Rendeiro said. “By combining histology images with artificial intelligence, we can detect patterns of biological aging that are invisible to the human eye.”
The models revealed that aging doesn’t progress uniformly across organs. Lungs and kidneys showed early signs of aging between 20 and 40. The uterus followed a different path, with periods of faster aging later in life.
This uneven pace indicates aging is shaped by both overall health and conditions specific to each organ. The findings show biological age isn’t a single measure for the whole body but varies by system.
From biopsies to blood tests
Since biopsies are invasive and impractical for routine monitoring, the team investigated whether blood tests could detect the same aging signals. They connected the tissue-based findings to gene expression profiles in blood, identifying patterns linked to diseases like Alzheimer’s, diabetes, and stroke.
Related: Smoking Cessation Struggles More in Developing Nations
“This is a conceptual leap,” said Iva Buljan, a doctoral researcher on the project. “We’re using the language of tissue aging, learned from images, and translating it into something readable from a routine blood draw.”
The approach could enable early detection of organ-specific decline if validated. Doctors might intervene before symptoms appear, and treatments could be tailored to how each patient’s organs age differently.
Ernesto Abila, a biomedical data scientist involved in the work, noted that deep learning helped interpret aging as “architectural remodeling, not just molecular drift.” The AI detected spatial patterns in tissue that reflect how cells reorganize over time—a dimension of aging often missed.
While still in early stages, the research suggests AI could transform medical diagnostics. Blood tests that reveal organ age might one day become as common as cholesterol checks, offering a clearer view of health than chronological age alone.
“Aging isn’t just about how many years you’ve lived,” said Yimin Zheng, another co-first author. “Different organs change at different speeds, influenced by both general health and local conditions.”
The study also highlights how blood-based methods could address complex health challenges beyond aging.
Leave a Reply