High-Quality Imaging Flow Cytometry Based on Acoustic Focusing and its Application in Label-Free Leukocyte Differential Coupled with Deep Neural Network

This paper presented a high-quality imaging flow cytometry based on acoustic focusing with demonstrative applications in label-free leukocyte differential coupled with deep neural network. Novelties of this study lied in 1) capture of high-quality cell images due to acoustic radiation forces, where all cells travelled along the standing wave node without deviation, enabling ∼100% capture of cells without losing focuses; 2) high classification accuracy due to deep neural network, where the best performance of leukocyte differential (99.4% at 5-part normal/abnormal leukocytes) was reported among label-free impedance and/or imaging flow cytometry.