AI Model Predicts Blood Stem Cell Quality by Tracking Live Cell Behavior

Published by University of Tokyo via Nature Communications

Researchers at the University of Tokyo have developed a novel system that predicts the quality of hematopoietic stem cells (HSCs) using real-time imaging and machine learning. By capturing continuous, 96-hour phase-contrast videos of single HSCs and analyzing their movement, division, and shape, the team was able to infer “stemness” via Hlf gene expression without damaging the cells. This time-based kinetic model outperforms traditional snapshot analysis and opens new pathways for improving safety and quality control in gene therapy and regenerative medicine.

Read the full article on Nature Communications:

👉 Quantitative phase imaging with temporal kinetics predicts hematopoietic stem cell diversity

Images used are for illustrative purposes only and do not depict actual persons, products, or facilities. For full detail, refer to the original article via the provided link.