Electron microscopy (EM) has long been a cornerstone in visualizing cellular structures, with volume EM (vEM) expanding this capability to three dimensions. However, limitations in imaging speed, quality, and sample size have persisted. Now, a novel solution called EMDiffuse, developed by researchers at the University of Hong Kong, promises to overcome these challenges using artificial intelligence. 

EMDiffuse, inspired by recent advancements in AI-powered image generation, employs diffusion model-based algorithms to enhance both 2D and 3D electron microscopy imaging. For conventional 2D EM, EMDiffuse restores high-quality visuals with exceptional ultrastructural details, even from noisy or low-resolution inputs. Its unique approach involves sampling solutions from target distributions, using low-quality images as constraints throughout the process.

Search Antibodies
Search Now Use our Antibody Search Tool to find the right antibody for your research. Filter
by Type, Application, Reactivity, Host, Clonality, Conjugate/Tag, and Isotype.

In the realm of vEM, EMDiffuse addresses the difficulty of capturing high-resolution 3D images of large samples, particularly in the depth dimension. The algorithm offers two flexible approaches: utilizing isotropic training data to enhance axial resolution, or employing self-supervised techniques to improve depth resolution without specialized training data.

The restored volumes produced by EMDiffuse demonstrate remarkable accuracy in studying ultrastructural details, such as mitochondrial cristae and interactions between mitochondria and the endoplasmic reticulum. These features are typically challenging to observe in original anisotropic volumes. 

EMDiffuse represents an important advancement in the imaging capabilities of both EM and vEM, enhancing image quality and axial resolution of the data produced. “With this foundation, we can envision further development and acceleration of the EMDiffuse algorithm, paving the way for in-depth investigations into the intricate subcellular nanoscale ultrastructure within large biological systems,’ said Professor Haibo Jiang, one of the corresponding authors of the Nature Communications paper.