Structured illumination microscopy (SIM), with the advantages of full-field imaging and low photo-damage, is one of the most well-established fluorescence super-resolution microscopy techniques that raised great interest in biological sciences. However, conventional SIM techniques generally require at least nine images for image reconstruction, and the quality of super-resolution significantly depends on high-accuracy illumination parameter estimation, which is usually computationally intense and time-consuming. To address these issues, we propose a robust seven-frame SIM reconstruction algorithm with accelerated correlation-enabled parameter estimation. First, a modulation-assigned spatial filter is employed to remove unreliable backgrounds associated with low signal-to-noise ratios. Then, we propose a coarse-to-fine accelerated correlation algorithm to eliminate the redundant iterations of the traditional correlation-based scheme. The frame reduction is achieved by a specially designed phase-shifting strategy combined with pixel-wise fluorescence pre-calibration. We experimentally demonstrate that, compared with conventional iterative correlation-based methods, the proposed algorithm improves the computational efficiency by a factor of 4.5 while maintaining high accuracy illumination parameter estimation. Meanwhile, our method achieves high-quality super-resolution reconstruction even with a reduction in two raw images, which improves the efficiency of image acquisition and ensures the robustness toward complex experimental environments.

Supplementary Material

You do not currently have access to this content.