摘要
成簇的微钙化点是早期乳腺癌的一个重要表征,用计算机辅助诊断时需要在乳腺片图像中判断有无钙化点簇。提出了一种利用Daubechies小波对乳腺图像进行增强,用二维熵阈值分割法对增强后的图像进行阈值分割,然后采用图论方法提取出成簇钙化点的方法。从而实现了计算机辅助诊断系统辅助医师判断恶性钙化。
Clustered microcalcifications is one of the main features of breast cancer. Whether there are clustered microcalcifications in mammograms should be estimated during computer-aided diagnosis. Image of mammograms are enhanced by Daubechies wavelet. Thresholding is implemented by 2D entropy method. Clustered rnicrocaldfications are detected with method of graphics theory. These make it possible for doctors to diagnose malignant calcications.
出处
《光学技术》
EI
CAS
CSCD
北大核心
2007年第5期769-771,共3页
Optical Technique
关键词
成簇微钙化点
小波变换
图论
乳腺癌
计算机辅助诊断
clustered rnicrocalcifications
wavelet transformation
graphics theory
breast cancer
CAD