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

Relevance vector machine for automatic detection of clustered microcalcifications.

Liyang Wei1, Yongyi Yang, Robert M Nishikawa

  • 1Department of Biomedical Engineering, Illinois Institute of Technology, Chicago, IL 60616, USA.

IEEE Transactions on Medical Imaging
|October 19, 2005
PubMed
Summary

This study introduces a machine learning approach using relevance vector machines (RVM) for detecting clustered microcalcifications (MCs) in mammograms. The RVM method significantly improves computational efficiency for breast cancer screening while maintaining high accuracy.

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

A upper critical solution temperature- and synergistic antioxidant-responsive tremella polysaccharide hydrogel for enhancing diabetic wound healing.

Biomaterials advances·2026
Same author

Ready, Set, Coach: A Mixed Methods Pilot Study Exploring the Impact of Implementation Coaching to Optimize the Allied Health Assistant Workforce in Public Health Settings.

Journal of multidisciplinary healthcare·2026
Same author

A long noncoding RNA modulates anthocyanin biosynthesis in Camellia sinensis.

Communications biology·2026
Same author

Differential physiological responses and transcriptome co-expression networks of salt-tolerant and salt-sensitive foxtail millet (<i>Setaria italica</i> (L.)) under salt stress.

Frontiers in plant science·2026
Same author

A data-driven respiratory motion correction for pediatric DMSA renal SPECT imaging: A simulation study.

Medical physics·2026
Same author

Mammo-GAN-Assisted Deep Network Training Scheme for Lesion Detection.

Journal of imaging informatics in medicine·2025

Area of Science:

  • Medical Imaging
  • Machine Learning
  • Breast Cancer Detection

Background:

  • Clustered microcalcifications (MCs) are critical early indicators of breast cancer in mammograms.
  • Accurate detection of MCs is essential for effective computer-aided detection (CADe) systems.

Purpose of the Study:

  • To propose and evaluate a novel machine learning technique, relevance vector machine (RVM), for detecting MCs in digital mammograms.
  • To develop computationally efficient algorithms for MC detection by leveraging the sparse properties of RVM.

Main Methods:

  • Formulated MC detection as a supervised learning problem, employing RVM as a classifier.
  • Developed a two-stage classification network utilizing a linear RVM for initial non-MC pixel elimination, enhancing speed.
  • Evaluated performance on 141 clinical mammograms using free-response receiver operating characteristic (FROC) curves and compared with support vector machine (SVM).

Related Experiment Videos

Main Results:

  • The RVM classifier demonstrated comparable accuracy to SVM while significantly reducing computational complexity.
  • The two-stage RVM approach reduced mammogram processing time from 250 seconds (SVM) to 7.26 seconds, a nearly 35-fold improvement.
  • RVM proved advantageous for real-time processing of MC clusters.

Conclusions:

  • Relevance vector machines offer an accurate and computationally efficient alternative for detecting microcalcifications in mammograms.
  • The proposed two-stage RVM method is highly suitable for real-time computer-aided detection systems in breast cancer screening.
  • RVM's sparse decision function property is key to its efficiency in medical image analysis tasks.