Signal Enhancement Ratios (SERs) in Breast Carcinomas Measured by 3D Contrast-MRI and Verified by Histopathology

Abstract

Our work in breast MRI has focused on an imaging technique and analysis method, directed toward defining the extent of malignant lesions in patients with confirmed breast carcinoma. We developed a 3-point contrast-MRI method to maximize anatomic (sensitivity) and biologic (specificity) information in a single exam. One data set is acquired at baseline (pre-contrast), S(sub 0); one early post-contrast, S(sub 1); and one late post-contrast, S(sub 2). The SER index, defined as (S(sub 1) - S(sub 0)) / (S(sub 2) - S(sub 0)), compares early to late enhancement. Preliminary studies suggested a relationship between SER value and tumor grade for invasive carcinomas. Our overall objective has been to develop and characterize this technique to be used with both high diagnostic and staging accuracy in evaluating the breast. Both the data acquisition and image analysis techniques are straightforward. We have aimed to reduce the computational complexity and to develop automated algorithms for analysis that can reduce inter- and intra-observer variability in making diagnostic and staging assessments. Our performance assessments to date demonstrate a 25% specificity improvement for this 3-time point method compared to a static' (2-point) method, approaching or exceeding specificity improvements reported with dynamic imaging techniques. Our staging results show that MRI performs substantially better than mammography in demonstrating disease extent, particularly in cases of multi-focal cancer and ductal carcinoma in situ. These results suggest that MRI may be cost-effective when used pre-surgically.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 1998
Accession Number
ADA366633

Entities

People

  • Nola Hylton

Organizations

  • University of California, San Francisco

Tags

DTIC Thesaurus Topics

  • Abstracts
  • Accuracy
  • Acquisition
  • Algorithms
  • Biomedical Research
  • Breast Cancer
  • Carcinoma
  • Data Acquisition
  • High Resolution
  • Histopathology
  • Imaging Techniques
  • Laboratory Animals
  • Magnetic Resonance
  • Mammography
  • Materials
  • Neoplasms
  • Sensitivity

Fields of Study

  • Medicine

Readers

  • Applied Combinatorial Optimization and Logic Circuit Design.
  • Medical Imaging.