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: Feb 28, 2026

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

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

Published on: August 30, 2013

43.7K

Improving Normal/Abnormal and Benign/Malignant Classifications in Mammography with ROI-Stratified Deep Learning.

Kenji Yoshitsugu1, Kazumasa Kishimoto2, Tadamasa Takemura1

  • 1Graduate School of Information Science, University of Hyogo, 7-1-28 Minatojima Minamimachi, Chuo-ku, Kobe-shi 650-0047, Hyogo, Japan.

Bioengineering (Basel, Switzerland)
|February 27, 2026
PubMed
Summary
This summary is machine-generated.

Stratifying mammography images by region of interest (ROI) presence improves deep learning (DL) classification accuracy for breast cancer screening. This ROI-based approach shows greater utility with larger datasets, enhancing diagnostic performance.

Keywords:
ROI-stratifieddeep learningmammographyregion of interest

More Related Videos

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

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

Published on: August 16, 2020

7.6K

Related Experiment Videos

Last Updated: Feb 28, 2026

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

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

Published on: August 30, 2013

43.7K
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

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

Published on: August 16, 2020

7.6K

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

Exploring Factors Influencing Nursing Task Prioritization for Supportive Information System Design: Qualitative Study With Thematic Analysis.

JMIR human factors·2026
Same author

Detection of Polyphonic Alarm Sounds From Medical Devices Using Frequency-Enhanced Deep Learning: Simulation Study.

JMIR medical informatics·2025
Same author

Improving Benign and Malignant Classifications in Mammography with ROI-Stratified Deep Learning.

Bioengineering (Basel, Switzerland)·2025
See all related articles

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Oncology

Background:

  • Deep Learning (DL) is increasingly used for medical image analysis and diagnosis.
  • Mammography is a key tool for breast cancer screening, with ongoing research to improve its accuracy.

Purpose of the Study:

  • To evaluate if stratifying mammography images based on the presence or absence of a Region of Interest (ROI) enhances classification accuracy.
  • To compare the diagnostic performance of DL models with and without ROI-based stratification.

Main Methods:

  • Trained multiple DL models (ResNet, EfficientNet, SwinTransformer, ConvNeXt, MobileNet) independently on subgroups of mammography images (with/without ROI).
  • Integrated results from subgroup predictions for final classification.
  • Utilized publicly available datasets: VinDr., CDD-CESM, and DMID.

Main Results:

  • ROI-based stratification improved classification accuracy for both normal-abnormal and benign-malignant cases.
  • The benefit of ROI stratification for diagnostic accuracy in mammography increased with larger dataset sizes.
  • Comparison with non-stratified predictions confirmed the enhanced utility of the ROI approach.

Conclusions:

  • Stratifying mammography images by ROI presence is a valuable technique for improving DL-based breast cancer screening accuracy.
  • The effectiveness of this stratification method is particularly pronounced in large-scale datasets.
  • This approach offers a promising strategy for enhancing diagnostic performance in mammography analysis.