Computerized Interpretation of Dynamic Breast MRI

Abstract

One of the most important obstacles of clinical application of breast MRI is the lack of standardization in terms of interpretation guidelines. The purpose of the proposed research is to develop computerized methods to take full advantage of the wealth information that dynamic MRI of the breast offers to improve methods for the diagnosis and prognosis of breast cancer, The research involves investigation of automatic methods for image artifacts correction, tumor segmentation, and extraction of computerized features that help distinguish between benign and malignant lesions. During the first year, we have collected and maintained a dynamic breast MRI database that is well suitable for the proposed research on computerized interpretation of breast MR images. We have developed computerized methods for correction of shading artifacts, tumor segmentation, and feature extraction. The results have shown that computerized analysis and interpretation methods have great promise in increasing the objectivity, efficiency, and accuracy of the diagnosis of breast cancer.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 01, 2004
Accession Number
ADA426322

Entities

People

  • Weijie Chen

Organizations

  • University of Chicago

Tags

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Breast Cancer
  • Computer Vision
  • Databases
  • Diagnostic Imaging
  • Discriminant Analysis
  • Electronic Mail
  • Feature Extraction
  • Health Services
  • Information Processing
  • Information Science
  • Magnetic Resonance
  • Medical Personnel
  • Neoplasms
  • Pattern Recognition
  • Three Dimensional

Fields of Study

  • Medicine

Readers

  • Computer Vision.
  • Oncology and Biomarker-Based Cancer Detection.
  • Systems Analysis and Design

Technology Areas

  • AI & ML