Kelvin C. M. Lee

CytoMAD improves label-free morphological analysis of cell heterogeneity across batches by distilling morphological information and enabling drug response prediction without manual annotation.

The paper introduces DIF, a method to focus polydisperse micro-particles without size-dependent dispersion, achieving high efficiency and uniformity in microfluidic applications.

PARC is a scalable graph-based clustering algorithm for large single-cell datasets, outperforming existing methods in speed and rare population detection.