Cell cycle stage classification using phase imaging with computational specificity

Yuchen R. He,Shenghua He,Mikhail E. Kandel,Young Jae Lee,Chenfei Hu,Nahil Sobh,Mark A. Anastasio,Gabriel Popescu
DOI: https://doi.org/10.1101/2021.11.05.467526
2021-11-07
Abstract:Abstract Traditional methods for cell cycle stage classification rely heavily on fluorescence microscopy to monitor nuclear dynamics. These methods inevitably face the typical phototoxicity and photobleaching limitations of fluorescence imaging. Here, we present a cell cycle detection workflow using the principle of phase imaging with computational specificity (PICS). The proposed method uses neural networks to extract cell cycle-dependent features from quantitative phase imaging (QPI) measurements directly. Our results indicate that this approach attains very good accuracy in classifying live cells into G1, S, and G2/M stages, respectively. We also demonstrate that the proposed method can be applied to study single-cell dynamics within the cell cycle as well as cell population distribution across different stages of the cell cycle. We envision that the proposed method can become a nondestructive tool to analyze cell cycle progression in fields ranging from cell biology to biopharma applications. Teaser We present a non-destructive, high-throughput method for cell cycle detection combining label-free imaging and deep learning.
What problem does this paper attempt to address?