A new computer-based tool has been developed to study ageing by analysing routine biopsy images. Instead of directly measuring DNA in a lab, this method looks at images of cells and tissues to detect subtle structural changes. These changes can reveal what is happening at the ends of chromosomes, where protective sections of DNA—called telomeres—are located. The findings were published in Cell Reports Methods on March 16, 2026.
The tool, called TLPath, was created by scientists at the Sanford Burnham Prebys Medical Discovery Institute. It is based on the idea that as cells age, their shape and structure change in ways that can be measured. By analysing these patterns in biopsy images, the model can estimate the length of telomeres without needing specialised laboratory tests.
Telomeres are repeating DNA sequences that protect the ends of chromosomes. Each time a cell divides, these telomeres become slightly shorter. This is a natural process, but over time, shorter telomeres are linked to ageing and a higher risk of diseases such as heart disease and cancer. Scientists have long known that telomere length is related to a person’s age, but measuring it directly can be complex, expensive, and difficult to scale for large studies.
To build the model, researchers used data from the Genotype-Tissue Expression Project, a large research programme supported by the National Institutes of Health. This dataset included thousands of high-quality biopsy images from different tissue types, along with laboratory measurements of telomere length. By training the model on this paired data, the system learned to recognise visual patterns linked to telomere length.
The model works by breaking each biopsy image into many small sections and analysing each one in detail. It looks for hundreds of structural features within cells and tissues, then combines this information to produce an overall prediction. After training, TLPath was able to accurately estimate telomere length in new samples it had not seen before. In fact, it performed better than simply using a person’s age to make predictions and could even detect differences between individuals of the same age.
This approach could make it much easier to study ageing and related diseases. Because biopsy slides are already commonly created in clinical care, the main requirement is to digitise and share them for research. With wider access to such data, tools like TLPath could help scientists better understand how ageing works and support efforts to improve health outcomes as people grow older.
More information: Anamika Yadav et al, Tissue morphology predicts telomere shortening in human tissues, Cell Reports Methods. DOI: 10.1016/j.crmeth.2026.101336
Journal information: Cell Reports Methods Provided by Sanford Burnham Prebys
