SVM and neural networks comparison in mammographic CAD

Annu Int Conf IEEE Eng Med Biol Soc. 2007:2007:3204-7. doi: 10.1109/IEMBS.2007.4353011.

Abstract

The purpose of this work is to compare the performance of Support Vector Machines (SVM) and Multi-Layer Perceptron (MLP) in the task of detection and diagnosis of microcalcification clusters in mammograms (MCCs). As data source, the "Digital Database for Screening Mammography" (DDSM) was used. The results show a similar performance for SVM and MLP, in both tasks, detection and diagnosis (slightly better for MLP in detection).

Publication types

  • Comparative Study
  • Evaluation Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Breast Diseases / diagnostic imaging*
  • Calcinosis / diagnostic imaging*
  • Expert Systems*
  • Female
  • Humans
  • Mammography / methods*
  • Neural Networks, Computer*
  • Pattern Recognition, Automated / methods*
  • Radiographic Image Enhancement / methods
  • Radiographic Image Interpretation, Computer-Assisted / methods*
  • Reproducibility of Results
  • Sensitivity and Specificity