In this research, the authors investigate the reconstruction algorithms in super-resolution structured illumination microscopy (SIM) and noise-specific artifacts that limit their applicability for lower signal-to-noise data, using high resolution fluorescent structures nanofabricated in the SEM. As a result of their study, they present a physically realistic noise model that explains the structured noise artifact. The methods, proposed by the authors, can help to enhance objectivity by eliminating ad hoc user-adjustable reconstruction parameters in favor of physical parameters.
Super-resolution structured illumination microscopy (SIM) is widely used for biological imaging, as it offers increased spatial resolution and optical sectioning with strongly enhanced contrast. However, current image reconstruction methods depend on ad hoc tuneable parameters and are susceptible to various types of artifacts. The purpose of the research was to design methods for optimizing contrast or a natural noise appearance, and to eliminate ad hoc reconstruction parameters, and, as a result, make the representation of objects in SIM images as objective as possible.
The authors of the paper propose three new, complementary image reconstruction methods for SIM. The authors recommend using the True-Wiener reconstruction as the default, as this method offers the best overall compromise between contrast, resolution, and noise profile. The proposed True-Wiener filtering method can also be applied to widefield imaging, scanning microscopy approaches and, be extended to lattice light sheet microscopy and to tomographic imaging modalities. The authors conclude that the use of spectral SNR and the generation of image representations with a flat-noise spectrum open up objective methods to assess the relative benefits of any super-resolution or deconvolution method.
Image source: Smith, C.S., Slotman, J.A., Schermelleh, L. et al. Structured illumination microscopy with noise-controlled image reconstructions. Nat Methods 18, 821–828 (2021). https://doi.org/10.1038/s41592-021-01167-7