Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Apr 30, 2026

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.6K

A Hybrid CNN-Transformer Deep Learning Model for Differentiating Benign and Malignant Breast Tumors Using Multi-View

Qi Zhang1, Ruizhuo Li1, Pan Tang1

  • 1Department of Medical Ultrasonics, The Second Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China.

Technology in Cancer Research & Treatment
|April 29, 2026
PubMed
Summary

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

HSPA6 Confers Lenvatinib Resistance in Hepatocellular Carcinoma by Promoting Phase Separation-Mediated TXNRD1 Stabilization to Suppress Ferroptosis.

Cancer research·2026
Same author

A Deep Learning Model Integrating Ultrasound Images and Multidimensional Clinical Information for Differentiating Benign and Malignant Non-Mass Lesions.

Diagnostics (Basel, Switzerland)·2026
Same author

Multimodal deep learning with missing data robustness for enhanced early diagnosis of coronary artery disease using CCTA, clinical, and ECG data.

BMC medical informatics and decision making·2026
Same author

Pre-clinical anti-tumor activity and preliminary safety of METTL3 inhibitors STM2457 and UZH1a in neuroblastoma.

Translational cancer research·2026
Same author

Adverse drug reactions among children with tuberculosis in China: a multicentre study, 2017-2022.

Annals of medicine·2026
Same author

Screening high-quality lignins extracted from biomass to prepare biobased wood adhesives.

Nature protocols·2026

A new deep learning system analyzes multiple breast ultrasound images to improve cancer diagnosis. This multi-view approach enhances accuracy and reliability compared to single-image analysis and human experts.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Oncology

Background:

  • Single-image analysis of breast lesions using ultrasound has limitations in diagnostic accuracy.
  • A computer-aided diagnosis (CAD) system is needed to emulate clinical multi-view diagnostic processes for improved reliability.

Purpose of the Study:

  • To develop and evaluate a novel hybrid deep learning model for breast cancer diagnosis using multi-view ultrasound images.
  • To improve the diagnostic accuracy and robustness of breast lesion analysis by integrating multi-view information.

Main Methods:

  • A hybrid deep learning model combining Convolutional Neural Network (CNN) and Transformer modules was proposed.
  • EfficientNetV2 extracted spatial features, and a Transformer encoder fused information from multiple unordered ultrasound images.
Keywords:
EfficientNetV2breast cancerdeep learningtransformerultrasound imaging

Related Experiment Videos

Last Updated: Apr 30, 2026

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.6K
  • The model was trained on internal (840 patients) and external (133 patients) datasets, with lesion-level partitioning and prospective validation (188 lesions).

    • Main Results:

    • The hybrid CNN-Transformer model achieved high performance: 0.960 accuracy, 0.967 sensitivity, and 0.9788 AUC on the internal test set.
    • Robust generalization was demonstrated on the external dataset (0.940 accuracy, 0.9730 AUC) and prospective validation (0.952 accuracy, 0.9801 AUC).
    • The model significantly outperformed a single-image baseline (0.881 accuracy) and a senior ultrasound physician (0.849 accuracy).

    Conclusions:

    • The proposed hybrid deep learning system effectively analyzes multi-view ultrasound images for breast cancer diagnosis.
    • Fusing multi-view information enhances diagnostic accuracy and robustness, aligning with clinical reasoning.
    • The system surpasses the performance of single-image analysis methods and human experts in breast lesion diagnosis.