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: Jan 10, 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

43.5K

Unsupervised Hybrid framework for ANomaly Detection (HAND)- applied to Screening Mammogram.

Zhemin Zhang, Bhavika Patel, Bhavik Patel

    IEEE Journal of Biomedical and Health Informatics
    |November 28, 2025
    PubMed
    Summary

    Out-of-distribution (OOD) detection in AI mammogram screening is improved by the novel HAND model. This method accurately identifies abnormal samples, enhancing AI generalization and ensuring reliable screening performance across diverse datasets.

    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

    Structure, assembly, and mechanism of the MmpL family of transporters.

    Microbiology and molecular biology reviews : MMBR·2026
    Same author

    Quantum Machine Learning for Biomedical Classification Problems: A Feasibility Study on Real Quantum Hardware.

    Annals of biomedical engineering·2026
    Same author

    CLEAR-AI: confounder-aware learning for equitable and accurate reasoning in AI for diagnosis.

    Journal of medical imaging (Bellingham, Wash.)·2026
    Same author

    Multimodal artificial intelligence models for radiology.

    BJR artificial intelligence·2026
    Same author

    Small subunit isoform diversity underlies structural heterogeneity in native plant Rubisco.

    Proceedings of the National Academy of Sciences of the United States of America·2026
    Same author

    Extreme and moderate temperatures and risk of hospitalizations for pulmonary hypertension: an 11-year time-series study in Shanghai, China.

    Frontiers in medicine·2026

    Area of Science:

    • Artificial Intelligence
    • Medical Imaging Analysis
    • Machine Learning for Healthcare

    Background:

    • Out-of-distribution (OOD) detection is crucial for AI generalization in mammogram screening.
    • Distribution shifts in external datasets can significantly degrade AI model performance.
    • Unsupervised generative learning is preferred for OOD detection due to limited prior knowledge of external data characteristics.

    Purpose of the Study:

    • To develop a novel approach for detecting OOD samples in mammogram screening.
    • To improve the robustness and generalization of AI models used in mammography.
    • To provide an automated solution for quality control in external mammogram datasets.

    Main Methods:

    • Developed a novel hybrid CNN-transformer backbone named HAND for OOD detection.

    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

    7.3K

    Related Experiment Videos

    Last Updated: Jan 10, 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

    43.5K
    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

    7.3K
  • Incorporated synthetic OOD samples and a latent space discriminator.
  • Applied gradient reversal to OOD reconstruction loss to enhance OOD differentiation.
  • Computed an anomaly score based on reconstruction and discriminator loss.

    • Main Results:

    • The HAND model demonstrated superior performance in OOD detection compared to encoder-based, GAN-based, and other hybrid CNN+transformer baselines.
    • The proposed method effectively distinguishes between in-distribution (ID) and OOD samples.
    • Achieved strong results on both internal (RSNA) and external (Mayo Clinic) mammogram datasets.

    Conclusions:

    • The HAND pipeline offers an automated and efficient computational solution for domain-specific quality checks in external screening mammograms.
    • This approach yields actionable insights without direct exposure to private medical imaging data.
    • The HAND model enhances the reliability of AI in mammogram screening by effectively handling OOD data.