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 Concept Videos

You might also read

Related Articles

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

Sort by
Same author

VITAGRAPH: building a knowledge graph for biologically relevant learning tasks.

Scientific data·2026
Same author

Diagnostic accuracy of contrast-enhanced mammography in evaluating breast cancer response to neoadjuvant therapy: a systematic review and meta-analysis.

La Radiologia medica·2026
Same author

Immunomodulatory effect of ultrasound-guided cryoablation in early breast cancer: pilot study on blood and surgical samples.

European radiology experimental·2025
Same author

Radiomics-based MRI models for predicting breast cancer axillary lymph node involvement in comparison with Node-RADS: a proof-of-concept study.

European radiology experimental·2025
Same author

Why clinical context and sustainability matters: the role of high-concentration contrast in CEM.

Insights into imaging·2025
Same author

Impact of Body Composition Changes on Treatment-Related Toxicities and Clinical Outcomes in HER2-Positive Metastatic Breast Cancer Patients Receiving Trastuzumab Deruxtecan.

Cancers·2025

Related Experiment Video

Updated: Sep 16, 2025

Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
15:48

Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging

Published on: December 15, 2014

22.6K

Deep Learning with Transfer Learning on Digital Breast Tomosynthesis: A Radiomics-Based Model for Predicting Breast

Francesca Galati1, Roberto Maroncelli1,2, Chiara De Nardo1

  • 1Department of Radiological, Oncological and Pathological Sciences, Sapienza University of Rome, Piazzale Aldo Moro 5, 00185 Rome, Italy.

Diagnostics (Basel, Switzerland)
|July 12, 2025
PubMed
Summary

Deep learning models show potential for classifying breast lesions on digital breast tomosynthesis (DBT) images, aiding in cancer detection. However, current models exhibit moderate performance, particularly in sensitivity, requiring further development for clinical use.

Keywords:
breast cancer risk stratificationdeep learningdigital breast tomosynthesisradiomicstransfer learning

More Related Videos

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

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

Published on: November 30, 2022

2.9K
Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

8.3K

Related Experiment Videos

Last Updated: Sep 16, 2025

Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
15:48

Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging

Published on: December 15, 2014

22.6K
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

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

Published on: November 30, 2022

2.9K
Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

8.3K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Oncology

Background:

  • Digital breast tomosynthesis (DBT) is crucial for breast cancer detection but faces challenges in interpretation time and variability.
  • Deep learning (DL) offers a potential solution to enhance DBT image analysis and support clinical decisions.

Purpose of the Study:

  • To develop and evaluate two DL models, ResNet50 and DenseNet201, using transfer learning for binary classification of breast lesions (benign vs. malignant) on DBT images.
  • To assess the models' performance in supporting clinical decision-making and risk stratification for breast cancer.

Main Methods:

  • A retrospective study included 184 patients with confirmed benign or malignant breast lesions.
  • Two CNN architectures, ResNet50 and DenseNet201, were trained using transfer learning from ImageNet weights.
  • A 10-fold cross-validation with ensemble voting was employed, and performance was measured using ROC-AUC, accuracy, sensitivity, specificity, PPV, and NPV.

Main Results:

  • The ResNet50 model achieved a ROC-AUC of 63%, accuracy of 60%, sensitivity of 39%, and specificity of 75%.
  • The DenseNet201 model reported a lower ROC-AUC of 55% and accuracy of 55%.
  • Both models showed high specificity but suboptimal sensitivity, suggesting potential for ruling out malignancy.

Conclusions:

  • Transfer learning-based DL models are feasible for breast lesion classification on DBT images.
  • The study highlights the potential and limitations of AI in breast imaging, with moderate overall performance.
  • Further research is necessary to improve model robustness and clinical applicability.