Laser & Optoelectronics Progress, Volume. 58, Issue 18, 1811007(2021)
Deep Learning Based Fluorescence Microscopy Imaging Technologies and Applications
Fig. 2. Confocal microscopy results based on deep learning. (a) Restored images using the UTOM algorithm[106], where the scale bar represents 50 μm; (b) super-resolution imaging using BPGAN[107], where the scale bars are 10 μm and 20 μm; (c) imaging on neuronal mitochondria using PSSR[99], where the scale bar represents 10 μm
Fig. 3. Imaging results of deep learning in light-sheet fluorescence microscopy. (a) Reconstruction result of zebrafish hearts using DVSR algorithm[91]; (b) neural/nuclear reconstruction of the Deep-SLAM algorithm in mouse brain[100], where the scale bar is 100 μm; (c) fluorescent microspheres/mouse brain slice images obtained using CBS-Deep algorithm[90]
Fig. 4. Schematic of light field fluorescence microscope[120]
Fig. 6. Structure and principle of SIM system[122]
Fig. 7. SIM imaging results based on deep learning.(a) Cross-modal imaging results from TIRF to TIRF-SIM[108]; (b) U-Net-SIM3 super-resolution imaging effect under extreme low-light conditions[102], where the scale bar is 1 μm; (c) 3_SIM imaging results[93]; (d) 5_NSIM imaging results[109]; (e) imaging results of DFCAN/DFGAN[110], where the scale bar is 2 μm
Fig. 8. Principle of STED microscopy imaging[122]
Fig. 10. Improved imaging performance of single molecular positioning microscope by the deep learning algorithm. (a) Reconstructed microtubule images using Deep-STORM[95], where the scale bar is 0.5 μm; (b) super-resolution three-dimensional imaging using DeepSTORM3D[96], where the scale bar is 5 μm; (c) volume tracking results of telomeres in living mouse embryonic fibroblast (MEF) cells using DeepSTORM3D[96], where the scale bar is 2 μm
Fig. 11. Deep learning cloud computing platform ZeroCostDL4Mic [141].(a) Schematic of ZeroCostDL4Mic; (b) workflow of ZeroCostDL4Mic; (c) bioimage analysis tasks implemented with the ZeroCostDL4Mic platform
Get Citation
Copy Citation Text
Haoyu Li, Liying Qu, Zijie Hua, Xinwei Wang, Weisong Zhao, Jian Liu. Deep Learning Based Fluorescence Microscopy Imaging Technologies and Applications[J]. Laser & Optoelectronics Progress, 2021, 58(18): 1811007
Category: Imaging Systems
Received: Jun. 1, 2021
Accepted: Aug. 9, 2021
Published Online: Sep. 3, 2021
The Author Email: Liu Jian (liujian@hit.edu.cn)