Search research articles

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Quantifying grain boundary deformation mechanisms in small-grained metals.

Nature·2025

Same author

Observing deformation in situ.

Nature materials·2024

Same author

pycotem: An open source toolbox for online crystal defect characterization from TEM imaging and diffraction.

Journal of microscopy·2020

Same author

A general and robust analytical method for interface normal determination in TEM.

Ultramicroscopy·2020

Same author

3D nanostructural characterisation of grain boundaries in atom probe data utilising machine learning methods.

PloS one·2019

Same author

Atomic scale analyses of {\bb Z}-module defects in an NiZr alloy.

Acta crystallographica. Section A, Foundations and advances·2018

Same journal

Efficient methods for wave propagation in electron microscopy.

Ultramicroscopy·2026

Same journal

Unsupervised deep image prior for sparse-view and limited-angle electron tomography.

Ultramicroscopy·2026

Same journal

Determination of the structure of the tertiary phase in the alloy Al<sub>10</sub>Mo<sub>10</sub>Nb<sub>10</sub>Ta<sub>10</sub>Ti<sub>30</sub>Zr<sub>30</sub> using convergent beam electron diffraction.

Ultramicroscopy·2026

Same journal

Predictive drift compensation of multi-frame STEM via live scan modification.

Ultramicroscopy·2026

Same journal

Deep PACBED: Multitask analysis of PACBED images using deep neural networks.

Ultramicroscopy·2026

Same journal

Guided progressive reconstructive imaging: A new quantization-based framework for low-dose, high-throughput and real-time analytical ptychography.

Ultramicroscopy·2026

See all related articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Feb 28, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Deep learning approaches for dislocation segmentation in TEM.

Assya Boughrara¹, Christine Viala¹, Laurent Dupuy²

¹Université de Toulouse, CNRS, CEMES, Toulouse, France.

Ultramicroscopy

|February 26, 2026

Summary

This summary is machine-generated.

Deep learning models, including semi-supervised learning, significantly improve dislocation segmentation in transmission electron microscopy (TEM) images, approaching expert performance. This facilitates faster analysis of material properties.

Keywords:

Deep learning Dislocation segmentation Supervised and semi-supervised learning

More Related Videos

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

Related Experiment Videos

Last Updated: Feb 28, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

Area of Science:

Materials Science
Computational Materials Science
Data Science

Background:

Dislocation dynamics critically influence the mechanical properties of alloys.
Analyzing dislocations using transmission electron microscopy (TEM) requires specialized knowledge and is time-intensive.

Purpose of the Study:

To develop deep learning approaches for automated dislocation segmentation in TEM images.
To enhance the efficiency and accuracy of dislocation analysis across diverse materials and imaging conditions.

Main Methods:

Implementation of fully supervised learning (FSL) and semi-supervised learning (SSL) using an encoder-decoder neural network with a boundary-type loss.
Utilizing a large in-house unlabeled dataset to enrich feature descriptions in SSL.
Exploring domain adaptation with physically-grounded synthetic images generated via simulation.

Main Results:

SSL approaches demonstrated improved evaluation metrics compared to FSL, nearing human expert performance.
Direct transfer of knowledge from synthetic to real images showed limited success.
Synthetic images proved beneficial for improving predictions in challenging imaging scenarios.

Conclusions:

Deep learning, particularly SSL, offers a powerful tool to automate and improve dislocation segmentation in TEM.
While synthetic data has limitations for direct knowledge transfer, it can aid in specific challenging imaging conditions.
Automated dislocation density measurement is a promising future application.