Introduction: Our research focuses on improving the analysis of electron microscope images, which can reveal intricate details of materials at the microscopic level. This capability is critical to advancing the field of materials science.

Methodology: We developed SEGDC-UNet, a specialized network designed to enhance the segmentation of particles in electron microscope images. SEGDC-UNet incorporates innovative techniques like the GELU activation function and channel attention mechanism to improve accuracy and efficiency.

Key Findings: In comparative experiments, SEGDC-UNet outperformed other models in metrics such as Dice coefficient, IoU, Pixel Accuracy, and Recall. These metrics quantify the model's ability to accurately identify and analyze microscopic structures.

Significance: Automating and improving the accuracy of electron microscope image analysis accelerates discoveries in material science.

Conclusion: Our study demonstrates SEGDC-UNet's effectiveness in advancing electron microscope image analysis, contributing to ongoing research and innovation in materials science. Future research could explore further applications and enhancements of SEGDC-UNet to address broader challenges in microscopy and material analysis.

3D printing has become a useful tool for fast, reproducible and accessible manufacturing of optical components and microscope hardware, but often is only used to part-build imaging setups or specimen holders. We have combined a 3D-printed microscope body with transparent 3D-printed lenses to produce the first fully 3D-printed microscope setup, which can be made using open-source designs and low-cost 3D printers. Each microscope takes only a few hours to manufacture and has a total manufacturing cost of £7.00, opposed to the thousands of pounds that a commercial microscope may cost. We tested the performance of our fully 3D-printed microscope using standard optical verification approaches, using microscopic rulers and resolution test targets to determine the magnification and resolution of the system. We also demonstrated the utility of our fully 3D-printed microscope to histology imaging, using commonly found clinical specimens such as blood smears and stained tissue sections. The 3D-printed microscope could resolve individual blood cells across a 1.7 mm field of view and detecting anatomical features in thin kidney tissue sections. These applications show great promise for diagnostic imaging in the field where users can easily design, manufacture and implement the low-cost microscope in low-resource settings.