Traditionally, flow cytometers have been used to analyze and sort cell populations and other particles applying laser-based fluorescence detection methods. However, the use of fluorescent dyes is associated with certain limitations that can lead to undesirable deviations in the test results. Furthermore, fluorescent markers are lacking- for example, in bacterial studies. Additionally, the assessment of morphological cell parameters by conventional flow cytometry is limited in scope. To address these challenges, this study introduces an innovative approach in flow cytometry analysis by integrating Multi-Angle Pulse Shape Flow Cytometry (MAPS-FC) with deep learning techniques, specifically using a deep autoencoder model. We used this technology for label-free classification of cell cycle phases as a proof of principle.
Devaraj et al. (Mon,) studied this question.