Tak Yung Man

This paper introduces DUCK-Net, a deep learning model for automated segmentation of Ductular Reactions in murine liver injury, showing high accuracy and correlation with spatial transcriptomic data.

The paper reports a method to reprogram primary human hepatocytes into proliferative pre-cHep cells with high expansion potential and liver repopulation capacity, offering an alternative to PHHs for drug testing and disease modeling.

The paper develops an in silico model to predict host responses to varying degrees of liver injury and validates it against experimental data in mice.