Digital Redepleted of Stimulated Emission Depletion Microscopy for Noise Reduction and Resolution Improvement

Abstract

Stimulated emission depletion microscopy (STED) achieves resolution beyond the diffraction limit by employing a donut-shaped depletion laser that selectively reduces fluorescence at the periphery of the excitation area. The imaging quality of STED microscopy is closely tied to minimizing the intermediate light from the ring-depletion laser. In this study, we introduce a method termed “digital redepleted STED,” which uses frequency domain filtering to generate an optimal donut profile by subtracting the “perfect donut” signal from the original STED data. This approach effectively reduces background noise and enhances the STED resolution. Through simulation experiments, we demonstrate that digitally redepleted STED doubled the resolution. This method is compatible with a wide range of biological samples and can be adapted for two-organelle-structure STED and 3D STED applications. We compare the performance of digitally redepleted STED with that of digitally enhanced STED (De STED) and deconvolution methods (STED Decon) in terms of the signal-to-background ratio (SBR) and resolution as evaluation metrics, and we find that our method doubled the resolution and SBR for different samples compared with origin STED. Our results indicate that digitally redepleted STED outperforms both De STED and STED Decon for complicated sample like mitochondria. We anticipate that the digitally redepleted STED will have broad applicability due to its enhanced resolution, improved SBR, and ease of implementation.