Search research 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

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

Quinoxaline Terpolymer-Controlled Miscibility With Oligomeric Acceptors for Over 20% Efficiency, Highly Stable and Stretchable Polymer Solar Cells.

Advanced materials (Deerfield Beach, Fla.)·2026

Same author

Ozone major blood ozonation in the management of dermatomyositis: a case report.

Frontiers in immunology·2026

Same author

Genetic architecture and major genes for tuber skin texture in potato.

Horticulture research·2026

Same author

Navigating Uncertainty in MRI Diagnosis: A Human-AI Collaborative Strategy for Stratifying Clinically Significant Prostate Cancer.

Journal of magnetic resonance imaging : JMRI·2026

Same author

DDMPI: Diffusion Denoising for Magnetic Particle Imaging at the Low Concentration.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

Same author

Nutrition-associated health levels in cancer patients based on the ICF: an expanded study of 300 cases.

Frontiers in nutrition·2026

Search research articles

Related Experiment Video

Updated: Jun 2, 2025

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

DIFLF: A domain-invariant features learning framework for single-source domain generalization in mammogram

Wanfang Xie¹, Zhenyu Liu², Litao Zhao¹

¹School of Engineering Medicine, Beihang University, Beijing 100191, PR China; Key Laboratory of Big Data-Based Precision Medicine (Beihang University), Ministry of Industry and Information Technology of the People's Republic of China, Beijing 100191, PR China.

Computer Methods and Programs in Biomedicine

|January 15, 2025

Summary

This summary is machine-generated.

This study introduces a novel framework for single-source domain generalization (SSDG) in deep learning models for breast cancer screening. The proposed method effectively reduces domain shifts, improving model performance on unseen datasets.

Keywords:

Breast cancer Content-style disentanglement module Deep learning Domain generalization Mammogram, Style-augmentation module

More Related Videos

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Published on: November 19, 2018

Related Experiment Videos

Last Updated: Jun 2, 2025

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Published on: November 19, 2018

Area of Science:

Artificial Intelligence
Medical Imaging Analysis
Machine Learning for Healthcare

Background:

Deep learning (DL) models for breast cancer screening face challenges in generalizing across different institutions due to domain shifts.
Single-source domain generalization (SSDG) aims to train a model on one dataset and apply it to multiple unseen datasets.
Alleviating domain shifts with only one source dataset is a significant challenge in clinical DL applications.

Purpose of the Study:

To propose a domain-invariant feature learning framework (DIFLF) for single-source domain generalization.
To enhance the robustness and generalizability of DL models in breast cancer screening.
To reduce the impact of domain shifts in mammogram classification tasks.

Main Methods:

Developed a domain-invariant features learning framework (DIFLF) incorporating a style-augmentation module (SAM) and a content-style disentanglement module (CSDM).
SAM utilizes color jitter transforms to increase feature diversity and reduce model overfitting.
CSDM employs feature disentanglement units to extract domain-invariant content features, minimizing the influence of domain-specific styles.

Main Results:

DIFLF demonstrated excellent performance in classifying mammograms on unseen datasets (PRI2, INbreast, MIAS) with varying feature distributions.
Achieved high accuracy and AUC scores, including 0.917 accuracy and 0.928 AUC on PRI2, 0.882 accuracy and 0.893 AUC on INbreast, and 0.767 accuracy and 0.710 AUC on MIAS.
The framework effectively mitigated the impact of domain shifts, even with significant differences in feature distributions.

Conclusions:

The proposed DIFLF effectively alleviates domain shifts using a single source dataset.
DIFLF achieves excellent mammogram classification performance on unseen datasets, including those with substantial feature distribution differences.
The framework shows promise for robust and generalizable DL applications in clinical breast cancer screening.