Real-time super-resolution structured illumination microscopy: current progress in joint space and frequency reconstruction.

Due to its low light dose, ultra-high imaging speed and super-resolution, structured illumination microscopy (SIM) is now widely used in various applications to study dynamic interactions between intracellular structures of living cells. However, real-time imaging remains challenging due to the limitations of existing reconstruction algorithms, such as spatial frequency domain conversion, iterative parameter estimation, and complex deconvolution. To achieve "what you see is what you get", the reconstruction algorithm of SIM needs to be improved. Here, we first briefly introduce the principle of super-resolution microscopy and SIM. Then, we comparatively review the algorithms for reconstructing super-resolution images in SR-SIM and discuss the advantages and disadvantages of each algorithm. To achieve real-time reconstruction, we propose a joint space and frequency reconstruction (JSFR) framework. Based on the JSFR framework, we realize a high image reconstruction speed and demonstrate its capability in real-time artifact reduction super-resolution imaging for 2D-SIM, 3D-SIM, and nonlinear SIM. Finally, we explore the prospects of the proposed technique by discussing its potential applications as a data platform for deep learning and live cell observation. .