Related Experiment Video
Updated: Aug 7, 2026
10:25
Deep Learning-Based Segmentation of Cryo-Electron Tomograms
Published on: November 11, 2022
Accessible deep learning for automated segmentation of supported nanoparticles in electron microscopy
Christian Vedel Petersen1, Marc Lindgaard1, Nikolaj Nguyen1
1Department of Energy Conversion and Storage, Technical University of Denmark 2800 Kgs. Lyngby Denmark psjq@dtu.dk yanhu@dtu.dk.
Nanoscale Advances
|August 6, 2026
Summary
This study introduces a deep learning model for fast and accurate nanoparticle segmentation in electron microscopy images. The accessible application enables efficient analysis, even with limited data and resources.
Area of Science:
- Materials Science
- Nanotechnology
- Catalysis Research
Background:
- Accurate nanoparticle size and morphology quantification is crucial for heterogeneous catalysis and energy conversion.
- Manual segmentation of nanoparticles in electron microscopy images is time-consuming, subjective, and difficult to scale.
Purpose of the Study:
- To develop an accessible and efficient deep learning model for automated nanoparticle segmentation.
- To create a functional analysis application for rapid segmentation and statistical quantification of nanoparticles.
Main Methods:
- Implementation of a deep learning model for image segmentation.
- Integration of the model into a user-friendly analysis application.
- Utilizing minimal annotated data and modest computational resources.
Main Results:
- Achieved fast and accurate segmentation of nanoparticles in electron microscopy images.
- Demonstrated the feasibility of high-quality segmentation with limited data.
- The developed application provides efficient statistical quantification.
Conclusions:
- Deep learning offers an efficient solution for nanoparticle segmentation in electron microscopy.
- The developed application democratizes advanced nanoparticle analysis.
- This approach accelerates research in catalysis and energy conversion.
