AI-driven label-free quantification of cell viability using live-cell analysis

Live-cell imaging enables acquisition of phase contrast images and provides an ideal platform to study multi-faceted biological paradigms in drug discovery. This is vital to our understanding of human diseases and treatment strategies. In this application note, Sartorius describes an automated, robust solution for label-free cell segmentation and live/dead classification of individual cells using integrated AI-based software. The Incucyte® AI Cell Health Analysis Software Module, driven by trained convolutional neural networks (CNN), allows for reliable monitoring of cell viability in a non-perturbing unbiased manner with minimal user input. Here, Sartorius shows validation of the analysis software across a wide range of live and dead adherent and non-adherent cell types and exemplifies how this approach can provide high-throughput, physiologically relevant insights into cell health through accurately predicting cell death across multiple treatments.