Label-free cell cycle stage classification using low-cost retrofittable quantitative phase imaging with machine learning
Poster presented as part of the Crick BioImage Analysis Symposium 2023.
Phototoxicity and photobleaching limit use of fluorescence-based cell-cycle reporters, e.g. PCNA, FUCCI, for long-term timelapse imaging. Quantitative phase imaging (QPI) using polarisation differential phase contrast microscopy (pDPC) offers label morphological information. We present a label-free approach using QPI implemented with polarisation differential contrast microscopy (pDPC), combined with machine learning for image feature extraction to classify cell cycle stage based on QPI images alone. Endogenously labelled PCNA offers an ‘all-in-one’ single channel fluorescent reporter of cell cycle stage to provide ground truth training data.
Permission has been given by authors to upload to Crick Figshare. Copyright remains with the original authors.