Correlative Feature Analysis for Multimodality Breast CAD

Abstract

The purpose of the study is to develop correlative feature analysis methods for integrating image information from multimodality breast images, taking advantage of the information from different views and/or different modalities, and thus improving the sensitivity and specificity of breast cancer diagnosis. Identifying the corresponding image pair of a lesion is an essential step for this purpose. During the past three years, we have built a multi-modality database which includes FFDM, breast US and DCE-MR images. We also developed computerized correlative feature analysis methods including automatic lesion segmentation, feature extraction and selection, feature correlation analysis and image pair classification in differentiating corresponding and non corresponding lesions across different mammographic views and/or different imaging modalities. The results show that the proposed correlative feature analysis is effective and robust for the discrimination between corresponding and non-corresponding lesion pairs.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2009
Accession Number
ADA517231

Entities

People

  • Yading Yuan

Organizations

  • University of Chicago

Tags

Communities of Interest

  • C4I

DTIC Thesaurus Topics

  • Breast Cancer
  • Computational Science
  • Computer Vision
  • Data Science
  • Databases
  • Feature Extraction
  • Health Services
  • Image Processing
  • Information Science
  • Machine Learning
  • Medical Personnel
  • Neoplasms
  • Neural Networks
  • Pattern Recognition
  • Statistics
  • Three Dimensional
  • United States

Fields of Study

  • Physics

Readers

  • Neural Network Machine Learning.
  • Oncology and Biomarker-Based Cancer Detection.

Technology Areas

  • AI & ML