Abstract
For high-throughput single cell analysis with a high accuracy, it is vital to develop a high-speed three-dimensional (3D) imaging method. For this endeavor, we propose a single-shot 3D cell imaging method that can achieve diffraction-limited spatial resolution and sub-millisecond temporal resolution. This method is realized through training a deep neural network (DNN) in an angle-multiplexed optical diffraction tomography (ODT) system to reconstruct the 3D refractive index maps of cells. Cells of various types are reconstructed in 3D using this method and the results are validated with a beam propagation-based reconstruction method. We applied this new imaging method for observing 3D red blood cell deformations in microfluidic channels and demonstrating 3D image flow cytometry at a throughput of around 5,000 cells/second. We envision this new cell tomography method will find a wide range of applications in biology and medicine. © 2021 The Author(s)
© 2021 The Author(s)
PDF ArticleMore Like This
Baoliang Ge, Mo Deng, Yanping He, Yijin Wang, Md Habibur Rahman, Yi-Ping Ho, Liting Duan, Zahid Yaqoob, George Barbastathis, Peter T. C. So, and Renjie Zhou
NF2C.4 Novel Techniques in Microscopy (NTM) 2021
Daniele Pirone, Daniele G. Sirico, Lisa Miccio, Vittorio Bianco, Amedeo Capozzoli, Claudio Curcio, Angelo Liseno, Pasquale Memmolo, and Pietro Ferraro
M4A.4 Digital Holography and Three-Dimensional Imaging (DH) 2022
Yanping He and Renjie Zhou
3Th5A.2 3D Image Acquisition and Display: Technology, Perception and Applications (3D) 2022