Automated reconstruction of whole-embryo cell lineages by learning from sparse annotations

Cell lineages in developing embryos are reconstructed from time-lapse microscopy images. We present a method to automatically identify and track nuclei in time-lapse microscopy recordings of entire developing embryos. The method combines deep learning and global optimization. On a mouse dataset, it reconstructs 75.8% of cell lineages spanning 1 h, as compared to 31.8% for the competing method. Our approach improves understanding of where and when cell fate decisions are made in developing embryos, tissues, and organs.

Automated summary

The method combines deep learning and global optimization to automatically identify and track nuclei in developing embryos, improving understanding of cell fate decisions.