Access to SPIE eBooks is limited to subscribing institutions. Access is not available as part of an individual subscription. However, books can be purchased on SPIE.Org
Chapter 8:
A Complex Wavelet-Based Feature Extraction System for Microcalcification Detection in Digital Mammograms
Abstract
Breast cancer is one of the leading causes of death among women worldwide. In the United States, it is the most-common form of cancer among women. Women in the U. S. have about a one-in-eight lifetime risk of developing invasive breast cancer. Early detection can increase the survival rate and treatment options. Regular mammographic screening programs for women at certain ages or in high-risk groups are necessary. Mammography is the most common and convenient procedure for detecting nonpalpable cancers. It is an inexpensive practice that is highly effective even when the size of the breast abnormality is minimal. Breast cancer can be divided into three categories: microcalcifications, masses, and architectural distortions. One of the early signs of breast cancer is the presence of microcalcification clusters (MCCs) in the mammogram. Because MCCs have small size, irregular shape, and low contrast, they are often missed or misinterpreted by physicians. Therefore, an automatic and reliable computer-aided diagnosis (CADx) system can be very useful in helping radiologists analyze mammographic lesions that may indicate the presence of cancer. The CADx system can detect and verify those mammogram images where possible cancer evidences have been developed; the images will then be sent to radiologists for final evaluation. Microcalcifications are tiny deposits of calcium, which appear as small bright spots in the mammogram. Microcalcifications are characterized by clusters, type, and distribution properties. Figure 8.1 shows two images of MCCs (a, b) and two images of a normal mammogram (c, d). Microcalcification image analysis and detection is an extremely challenging task for the following three reasons: First of all, there is a large variability in the appearance of abnormalities. Likewise, abnormalities are often occluded or hidden in dense breast tissue. Perhaps most importantly, a CADx system for MCC detection is used in serious human disease detection; therefore, a need for near-perfect accuracy is required.
Online access to SPIE eBooks is limited to subscribing institutions.
CHAPTER 8
17 PAGES


SHARE
Back to Top