Advancing our understanding of life at the cellular and subcellular levels relies critically on the ability to visualize fine biological structures with high spatial resolution. In modern life science research, super-resolution fluorescence microscopy has become an indispensable tool, enabling imaging beyond the diffraction limit of conventional optical systems. Techniques such as structured illumination microscopy (SIM) offer a compelling balance between spatial resolution, imaging speed, and low phototoxicity, making them particularly suitable for long-term live-cell observation.
However, despite their widespread utility, existing super-resolution microscopy systems often suffer from significant practical limitations. High-performance implementations--especially interference-based SIM--typically require complex optical configurations, precise alignment, and bulky instrumentation, leading to high system cost and limited accessibility. These constraints have hindered the broader deployment of super-resolution imaging technologies in routine biological research and resource-limited environments.
Recently, a research team from the Smart Computational Imaging Laboratory (SCILab) at Nanjing University of Science and Technology, led by Professor Chao Zuo, has developed a novel super-resolution imaging framework termed PCA-iSIM. This approach enables high-resolution, real-time imaging on a compact and cost-effective platform, addressing key bottlenecks in conventional SIM systems.
Structured illumination microscopy achieves super-resolution by encoding high-frequency sample information into the passband of the optical transfer function through patterned illumination. Among various implementations, digital micromirror device (DMD)-based incoherent SIM systems have attracted attention due to their compactness and low hardware complexity. However, a long-standing challenge has limited their performance: under high-frequency structured illumination, the modulation depth of the projected fringes is severely attenuated by the system’s optical transfer function. As a result, fringe visibility becomes extremely weak, making it difficult to accurately estimate illumination parameters and reconstruct high-fidelity super-resolution images. This limitation has long posed a major barrier to achieving super-resolution in DMD-based SIM systems.
To overcome this fundamental limitation, the researchers proposed a new computational imaging strategy that integrates high-modulation coefficient mapping with principal component analysis (PCA). Instead of directly relying on weak high-frequency modulation signals, the method establishes a mapping relationship between low-frequency, high-visibility illumination patterns and their corresponding high-frequency counterparts. This enables reliable estimation of illumination wave vectors even under conditions of severely reduced modulation contrast.
Building upon this mapping, PCA is employed to extract the dominant signal components from noisy measurements. Explaining the approach, Prof. Zuo says, “By leveraging PCA’s capability for dimensionality reduction and noise suppression, we are able to isolate the illumination-related information embedded in the data and accurately recover sub-pixel illumination parameters.” This combined framework significantly enhances the robustness and precision of parameter estimation, which is critical for achieving high-quality super-resolution reconstruction.
Using a custom-built DMD-based incoherent SIM system, the team experimentally validated the performance of PCA-iSIM. The results demonstrate a resolution enhancement exceeding 1.9×, corresponding to an effective spatial resolution of approximately 100 nm, while maintaining real-time reconstruction at speeds up to 30 frames per second. Notably, the system achieves these improvements with a substantially simplified optical design, reducing system complexity by nearly 70% compared to traditional laser-based SIM setups.
Beyond resolution enhancement, the proposed method also enables stable imaging under low signal-to-noise conditions and dynamic experimental environments. The researchers further demonstrated its capability for real-time observation of mitochondrial dynamics in living cells, revealing sub-diffraction structural changes that are otherwise inaccessible with conventional wide-field microscopy.
“This work fundamentally extends the performance ceiling of incoherent structured illumination microscopy,” the authors note. “By combining compact hardware with advanced computational reconstruction, PCA-iSIM opens new opportunities for accessible, high-performance super-resolution imaging.”
Laser & Photonics Review
Imaging analysis
5-Feb-2026
The authors declare that they have no competing interests.