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 Videos

Computer-aided detection in screening mammography: variability in cues.

Jay A Baker1, Joseph Y Lo, David M Delong

  • 1Departments of Radiology and Biomedical Engineering, Duke University Medical Center, Erwin Rd, Durham, NC 27710, USA. jay.baker@duke.edu.

Radiology
|September 11, 2004
PubMed
Summary

Computer-aided detection (CAD) systems show variability in marking true-positive breast cancer cues, with greater inconsistency for false positives. However, CAD systems remain consistent in the overall number of cancers detected across multiple runs.

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

Learning to Diagnose Privately: DP-Powered LLMs for Radiology Report Classification.

IEEE access : practical innovations, open solutions·2026
Same author

Breast area affects the performance of a commercial artificial intelligence algorithm assessment of negative digital breast tomosynthesis exams.

European journal of radiology·2026
Same author

Demographic distribution matching between real-world and virtual phantom population.

Medical physics·2026
Same author

Utility of the virtual imaging trials methodology for objective characterization of AI systems and training data.

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

Cross-National Radiomics Validation Using Mammography to Predict Occult Invasion in Ductal Carcinoma in Situ.

Radiology·2025
Same author

Evaluation of unified harmonization of CT images across multiple tasks: A step towards AI generalizability.

Medical physics·2025

Area of Science:

  • Radiology
  • Medical Imaging
  • Artificial Intelligence in Medicine

Background:

  • Computer-aided detection (CAD) systems are increasingly used in mammography to assist radiologists.
  • Evaluating the reliability and consistency of CAD systems is crucial for their effective clinical implementation.
  • Variability in CAD performance can impact diagnostic accuracy and workflow.

Purpose of the Study:

  • To assess the variability of true-positive and false-positive marks generated by a commercial computer-aided detection (CAD) system.
  • To evaluate the reproducibility of CAD analysis for breast malignancies detected during screening mammography.
  • To determine case-based and image-based sensitivity and reproducibility of the CAD system.

Main Methods:

  • Fifty breast cancers from a screening population were analyzed using a commercial CAD system.

Related Experiment Videos

  • Mammograms (craniocaudal and mediolateral oblique views) were digitized and analyzed 10 times by the CAD system.
  • A radiologist compared CAD output to malignancy location, assessing accuracy in one or both views, or neither. Sensitivity and reproducibility were calculated.

    • Main Results:

    • Overall case-based sensitivity was 82.4%, while image-based sensitivity was 61.1%.
    • Variability in true-positive CAD cues occurred in 28% of cases and 33% of mammographic views.
    • False-positive marks exhibited significantly greater variability than true-positive marks, though overall cancer detection remained consistent (40-43/50 cancers per run).

    Conclusions:

    • The study demonstrated inconsistency in CAD analysis for screening-detected breast cancers.
    • Despite variability in specific cue marking, the CAD system showed reasonable consistency in the total number of cancers identified across repeated analyses.
    • The CAD system exhibited higher variability for false-positive marks compared to true-positive marks.