Computerized Analysis of MR and Ultrasound Images of Breast Lesions

Abstract

We are developing methods for the classification of lesions in ultrasound and MR images of the breast. We are retrospectively collecting large datasets of ultrasound and MR images of the breast with solid pathology and radiologists ratings. These cases include malignant lesions, benign lesions, and complex cysts. We are developing noninvasive computerized methods for characterizing the lesions to yield an output related to the probability of malignancy. We plan to evaluate the efficacies of the new image analysis methods in the task of distinguishing between malignant and benign lesions. It is expected that the results from this research will aid radiologists in determining the likelihood of malignancy and in reducing the number of benign cases sent to biopsy.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 1999
Accession Number
ADA384095

Entities

People

  • Maryellen Lissak Giger

Organizations

  • University of Chicago

Tags

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Breast Cancer
  • Carcinoma
  • Computer Vision
  • Computers
  • Data Sets
  • Databases
  • Electronic Mail
  • Image Processing
  • Information Science
  • Magnetic Resonance
  • Medical Personnel
  • Neoplasms
  • Neural Networks
  • Three Dimensional
  • Two Dimensional
  • Ultrasounds

Fields of Study

  • Medicine

Readers

  • Medical Imaging.
  • Oncology and Biomarker-Based Cancer Detection.