Wavelet Representations for Digital Mammography.
Abstract
The goal of this project is to develop an interactive diagnostic tool for radiologists that shall refine the perception of mammographic features (including lesions, masses and calcifications). By improving the visualization of breast pathology we can increase the chances of early detection of breast cancers (improve quality) while requiring less time to evaluate mammograms for most patients (lower costs). We are investigating a methodology for accomplishing mammographic feature analysis through multiresolution representations. We show that overcomplete representations may be used to detect and enhance specific mammographic features within a continuum of scale space. Such focused reconstructions can allow a radiologist to examine interactively diagnostic features within a selected scale space. The improved contrast of mammographic features make these techniques appealing for computed aided diagnosis (CAD) and screening mammography. Screening mammography examinations are certain to grow substantially in the next few years, and analytic methods that can assist general radiologists in reading mammograms shall be of great importance.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1997
- Accession Number
- ADA349968
Entities
People
- Andrew F. Laine
Organizations
- University of Florida