The CAD Method for Microcalcification Detection: Independent of Sensor and Resolution
Abstract
The complishments of the first year's task in this project are summarized as follows: (1). Data base collection and truth file establishment for three sensors (2). Breast area segmentation for the mammograms (3). Basic algorithm design including nonlinear bank filter design, implementation and segmentation algorithm design and testing. These achievements provide an very important background for the future implementation of our long term aims of the project, which is the development of a more generalizable, automatic, and robust computer assisted diagnostic (CAD) method, for microcalcification cluster (MCC) detection using digital mammography; suitable for both digitized film and direct X-ray sensors. The specific aim of this proposal is to evaluate the feasibility of developing an entirely new class of adaptive multiresolution CAD method, implemented on non-linear filter banks, and compare this method to a wavelet transform method, implemented on linear filter banks, as previously reported.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 2000
- Accession Number
- ADA385946
Entities
People
- Wei Qian
Organizations
- University of South Florida