Enhanced detection of fluorescence fluctuations for high-throughput super-resolution imaging

• Published in: Nature photonics 2023-09, Vol.17 (9), p.806-813
• Main Authors: Zhao, Weisong, Zhao, Shiqun, Han, Zhenqian, Ding, Xiangyan, Hu, Guangwei, Qu, Liying, Huang, Yuanyuan, Wang, Xinwei, Mao, Heng, Jiu, Yaming, Hu, Ying, Tan, Jiubin, Ding, Xumin, Chen, Liangyi, Guo, Changliang, Li, Haoyu
• Format: Article
• Language: English
• Subjects: 631/57; 639/624/1107/328/2238; 639/624/1111/55; 639/766/930/328; Accuracy; Applied and Technical Physics; Autocorrelation; Biomedical materials; Deconvolution; Fluorescence; Frames; Image acquisition; Image resolution; Localization; Microscopy; Optical components; Optics; Photonics; Physics; Physics and Astronomy; Quantum Physics; Sparsity; Spatial discrimination; Spatial resolution; Visual field
• ISSN: 1749-4885; 1749-4893
• DOI: 10.1038/s41566-023-01234-9

High-throughput super-resolution (SR) imaging is attractive for rapid and high-precision profiling in a wide range of biomedical applications. However, current SR methods require sophisticated acquisition optics and long integration times to acquire a single field of view. By exploiting the natural photophysics of fluorescence, fluctuation-based microscopy techniques can routinely break the diffraction limit without requiring additional optical components. However, their long acquisition time still poses a challenge for high-throughput imaging and the visualization of transient cellular dynamics. Here we propose super-resolution imaging based on autocorrelation with two-step deconvolution (SACD). Our method notably reduces the number of frames required by maximizing the detectable fluorescence fluctuation behaviour in each measurement. SACD requires only 20 frames to achieve a twofold improvement in lateral and axial resolution, whereas current SR optical fluctuation imaging (SOFI) needs more than 1,000 frames. With an acquisition time of ~10 min, we record SR images with 128-nm resolution over a field of view of 2.0 mm × 1.4 mm, which includes more than 2,000 cells. By applying the continuity and sparsity joint constraint, sparse deconvolution-assisted SACD enables four-dimensional imaging of live cells and events such as mitochondrial fission and fusion. Overall, as an open-sourced module, we anticipate that SACD will improve accessibility to SR imaging, thus facilitating biological studies of cells and organisms with high throughput and low cost. Super-resolution imaging based on autocorrelation with two-step deconvolution (SACD) enables recording super-resolution images with 128-nm spatial resolution over a field of view of 2.0 mm × 1.4 mm within a 10-min acquisition time.