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

Classification of clustered microcalcifications using a Shape Cognitron neural network.

San Kan Lee1, Pau choo Chung, Chein I Chang

  • 1Department of Radiology, Taichung Veterans General Hospital, VACRS, 40705, Taichung, Taiwan, ROC.

Neural Networks : the Official Journal of the International Neural Network Society
|February 11, 2003
PubMed
Summary

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

Breast cancer detection of large size to DCIS by hypoxia and angiogenesis using NIRS.

Advances in experimental medicine and biology·2013
Same author

Comprehensive evaluation of CT pulmonary angiography for patients suspected of having pulmonary embolism.

The international journal of cardiovascular imaging·2009
Same author

Usefulness of multidetector-row computed tomography in evaluating adult cor triatriatum.

Texas Heart Institute journal·2008
Same author

Comprehensive evaluation of ischemic heart disease using MDCT.

AJR. American journal of roentgenology·2008
Same author

Neonatal cardiac multidetector row CT: why and how we do it.

Pediatric radiology·2008
Same author

Contrast enhancement in cardiac MDCT: comparison of iodixanol 320 versus iohexol 350.

AJR. American journal of roentgenology·2007

A novel Shape Cognitron (S-Cognitron) neural network effectively classifies clustered microcalcifications using universal feature planes. This shape recognition system achieves high sensitivity and specificity in mammogram analysis.

Area of Science:

  • Artificial Intelligence
  • Medical Imaging Analysis
  • Biomedical Engineering

Background:

  • Accurate classification of clustered microcalcifications is crucial for early breast cancer detection.
  • Existing methods may face challenges in capturing complex shape features for microcalcification analysis.

Purpose of the Study:

  • To introduce and evaluate a new shape recognition-based neural network, the Shape Cognitron (S-Cognitron), for classifying clustered microcalcifications.
  • To assess the performance of S-Cognitron using a standard mammogram database.

Main Methods:

  • Developed the Shape Cognitron (S-Cognitron) architecture with two modules and an intermediate 3D figure layer.
  • The first module includes shape orientation and complex layers for low- and second-order feature extraction.

Related Experiment Videos

  • The 3D figure layer extracts and displays shape curvatures, followed by a feature formation and probabilistic neural network classification layer.

    • Main Results:

    • The S-Cognitron system demonstrated promising performance on the Nijmegen mammogram database.
    • Achieved a sensitivity of 86.1% and a specificity of 74.1% in classifying clustered microcalcifications.

    Conclusions:

    • The Shape Cognitron (S-Cognitron) presents a viable approach for automated microcalcification classification.
    • The proposed architecture effectively utilizes universal feature planes for enhanced shape recognition in medical images.