Understanding the function of each individual cell and its differences in healthy and diseased individuals requires the analysis of large numbers of single cells. Researchers at the Technical University of Munich (TUM), including Professor Fabian Theis and his team, as well as Helmholtz Munich, have now made a significant advancement in this field: With a new method, millions of single cells can be analyzed efficiently and precisely using machine learning.
Self-supervised learning combines two approaches for this purpose: masked learning and contrastive learning. This methodology is used, for example, to analyze tissues based on individual cells and determine their different functions.
These advancements open up new possibilities for studying cellular changes in detail.
For more details, please refer to the original article on bionity.com.
