We demonstrate that a deep neural network can be trained to virtually refocus a two-dimensional fluorescence image onto user-defined three-dimensional (3D) surfaces within the sample. Using this method, termed Deep-Z, we imaged the neuronal activity of a Caenorhabditis elegans worm in 3D using a time sequence of fluorescence images acquired at a single focal plane, digitally increasing the depth-of-field by 20-fold without any axial scanning, additional hardware or a trade-off of imaging resolution and speed. Furthermore, we demonstrate that this approach can correct for sample drift, tilt and other aberrations, all digitally performed after the acquisition of a single fluorescence image. This framework also cross-connects different imaging modalities to each other, enabling 3D refocusing of a single wide-field fluorescence image to match confocal microscopy images acquired at different sample planes. Deep-Z has the potential to improve volumetric imaging speed while reducing challenges relating to sample drift, aberration and defocusing that are associated with standard 3D fluorescence microscopy.
Bibliographical noteFunding Information:
The authors acknowledge Y. Luo, X. Tong, T. Liu, H. C. Koydemir and Z. S. Ballard of University of California, Los Angeles (UCLA), as well as Leica Microsystems for their help with some of the experiments. The Ozcan group at UCLA acknowledges the support of Koc Group, National Science Foundation and the Howard Hughes Medical Institute. Y.W. also acknowledges the support of a SPIE John Kiel scholarship. Some of the reported optical microscopy experiments were performed at the Advanced Light Microscopy/Spectroscopy Laboratory and the Leica Microsystems Center of Excellence at the California NanoSystems Institute at UCLA with funding support from National Institutes of Health Shared Instrumentation grant S10OD025017 and National Science Foundation Major Research Instrumentation grant CHE-0722519. We also thank Double Helix Optics for providing their SPINDLE system and DH-PSF phase mask, which was used for engineered PSF data capture, and acknowledge X. Yang and M.P. Lake for their assistance with these engineered PSF experiments and related analysis.
© 2019, The Author(s), under exclusive licence to Springer Nature America, Inc.