Field-dependent deep learning enables high-throughput whole-cell 3D super-resolution imaging

Shuang Fu; Wei Shi; Tingdan Luo; Yingchuan He; Lulu Zhou; Jie Yang; Zhichao Yang; Jiadong Liu; Xiaotian Liu; Zhiyong Guo; Chengyu Yang; Chao Liu; Zhen-Li Huang; Jonas Ries; Mingjie Zhang; Peng Xi; Dayong Jin; Yiming Li

doi:10.1038/s41592-023-01775-5

Field-dependent deep learning enables high-throughput whole-cell 3D super-resolution imaging

Nat Methods. 2023 Mar;20(3):459-468. doi: 10.1038/s41592-023-01775-5. Epub 2023 Feb 23.

Authors

Shuang Fu^#¹, Wei Shi^#¹, Tingdan Luo¹, Yingchuan He¹, Lulu Zhou¹, Jie Yang¹, Zhichao Yang¹, Jiadong Liu², Xiaotian Liu², Zhiyong Guo¹, Chengyu Yang¹, Chao Liu¹, Zhen-Li Huang³, Jonas Ries⁴, Mingjie Zhang², Peng Xi^{1

5}, Dayong Jin^{1

6}, Yiming Li⁷

Affiliations

¹ Guangdong Provincial Key Laboratory of Advanced Biomaterials, Department of Biomedical Engineering, Southern University of Science and Technology, Shenzhen, China.
² School of Life Sciences, Southern University of Science and Technology, Shenzhen, China.
³ Key Laboratory of Biomedical Engineering of Hainan Province, School of Biomedical Engineering, Hainan University, Haikou, China.
⁴ European Molecular Biology Laboratory, Cell Biology and Biophysics, Heidelberg, Germany.
⁵ Department of Biomedical Engineering, College of Future Technology, Peking University, Beijing, China.
⁶ Institute for Biomedical Materials and Devices (IBMD), Faculty of Science, University of Technology Sydney, Sydney, New South Wales, Australia.
⁷ Guangdong Provincial Key Laboratory of Advanced Biomaterials, Department of Biomedical Engineering, Southern University of Science and Technology, Shenzhen, China. liym2019@sustech.edu.cn.

^# Contributed equally.

PMID: 36823335
DOI: 10.1038/s41592-023-01775-5

Abstract

Single-molecule localization microscopy in a typical wide-field setup has been widely used for investigating subcellular structures with super resolution; however, field-dependent aberrations restrict the field of view (FOV) to only tens of micrometers. Here, we present a deep-learning method for precise localization of spatially variant point emitters (FD-DeepLoc) over a large FOV covering the full chip of a modern sCMOS camera. Using a graphic processing unit-based vectorial point spread function (PSF) fitter, we can fast and accurately model the spatially variant PSF of a high numerical aperture objective in the entire FOV. Combined with deformable mirror-based optimal PSF engineering, we demonstrate high-accuracy three-dimensional single-molecule localization microscopy over a volume of ~180 × 180 × 5 μm³, allowing us to image mitochondria and nuclear pore complexes in entire cells in a single imaging cycle without hardware scanning; a 100-fold increase in throughput compared to the state of the art.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Deep Learning*
Imaging, Three-Dimensional / methods
Single Molecule Imaging / methods