AI detects signs of aging in blood stem cells from nuclear images
Aging progressively affects how our bodies function. Among other effects, it reduces the ability of the hematopoietic system—the organs and tissues responsible for producing blood cells—to maintain adequate blood cell production. Understanding and measuring this process is especially important for studying how blood stem cells age and identifying ways to preserve or restore their function.
A recent study has developed an artificial intelligence tool called ChromAgeNet that can identify signs of aging in hematopoietic stem cells using nuclear images. Hematopoietic stem cells, which reside in the organs and tissues responsible for blood cell production, lose their ability to maintain adequate blood cell production as people age.
Understanding and measuring this process is crucial for investigating cellular aging and finding ways to preserve or restore stem cell function. The researchers used a convolutional neural network to analyze 3D images of mouse hematopoietic stem cell nuclei stained with DAPI. This technique allowed ChromAgeNet to differentiate young cells from aged cells with 77% accuracy, outperforming a previous machine learning model based on chromatin features.
The model identified chromatin entropy, heterochromatin at the nucleus' periphery, and certain chromatin condensates as features predictive of an age-associated state. These findings suggest that artificial intelligence can help detect age-related changes in blood stem cells, providing valuable insights into cellular aging. The researchers also explored the model's potential as a tool for screening treatments that could modify age-associated characteristics.
While the study did not demonstrate functional rejuvenation of cells, it showed the potential of ChromAgeNet to detect age-associated changes in response to interventions and accelerate the identification of candidates for cell rejuvenation strategies.
Written by urgent.news from Medical Xpress's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
