Abstract
Structured illumination microscopy (SIM) has become a widely used tool for insight into biomedical challenges due to its rapid, long-term, and super-resolution (SR) imaging. However, artifacts that often appear in SIM images have long brought into question its fidelity, and might cause misinterpretation of biological structures. We present HiFi-SIM, a high-fidelity SIM reconstruction algorithm, by engineering the effective point spread function (PSF) into an ideal form. HiFi-SIM can effectively reduce commonly seen artifacts without loss of fine structures and improve the axial sectioning for samples with strong background. In particular, HiFi-SIM is not sensitive to the commonly used PSF and reconstruction parameters; hence, it lowers the requirements for dedicated PSF calibration and complicated parameter adjustment, thus promoting SIM as a daily imaging tool.
Highlights
Super-resolution structured illumination microscopy (SR-Structured illumination microscopy (SIM)), which breaks the optical diffraction barrier, offers an unprecedented opportunity for investigating biological structures at a ~100 nm scale[1,2]
The cortical endoplasmic reticulum (ER) network near the basal cell cortex were imaged by GI-SIM at 97 nm resolution and 266 Hz framerate over thousands of time points[9]
The reconstruction algorithms used by most commercial SIM setups and successful open-source packages, such as SIMToolbox[30], fairSIM31, and OpenSIM32, were based on the Wiener deconvolution procedure established by Heintzmann[33] and Gustafsson[34]
Summary
Super-resolution structured illumination microscopy (SR-SIM), which breaks the optical diffraction barrier, offers an unprecedented opportunity for investigating biological structures at a ~100 nm scale[1,2]. SR-SIM at video frame-rate, with a delay of less than 250 ms between acquisition and reconstruction[11] Despite these advances, the fidelity and quantification of SR-SIM are often challenged[12,13] because the final SR images heavily rely on post-processing algorithms that are prone to reconstruction artifacts[8,14,15]. Several studies have been conducted on reconstruction algorithms, including accurate illumination parameter estimation[20,21,22,23,24], iterative deconvolution[8,25,26], and fine tuning of reconstruction parameters[27] Despite all these efforts, artifacts that limit the implementation of SIM as a daily imaging tool still frequently appear in SIM images. Deep learning has shown great potential for SR-SIM reconstruction[28,29], but the results are closely related to SR images obtained via other methods for training neural network, the fidelity is still questioned
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
