摘要
研究了基于灰度空域统计特征以及灰度共生矩阵的医学乳腺X线图像的特征提取方法,以及这些特征对于数据挖掘中的两种算法——基于神经网络的算法和基于关联规则挖掘的算法在乳腺肿瘤检查和分类中的作用,结果表明这些特征在两种分类方法中均表现良好,对良性与恶性肿瘤分类的准确率均超过了75%.实验证明所提出的特征提取方法对于神经网络和关联规则的挖掘在乳腺X线图像分类中是有效的.
This paper investigate three types of texture features.stastical descriptors and stastical feature from decomposition of image on wavlet and gray level co-occurrence matrix, and the use of different data mining techniques, neural networks and association rule mining in anomaly detection and classification. The results show that the two approaches performed well,obtaining a classification accuracy reaching over 75% percent for both techniques. The experiments we conducted demonstrate the use and effctiveness of data mining method based on the features we extracted from the digital mammography.
出处
《陕西科技大学学报（自然科学版）》
2007年第1期117-120,共4页
Journal of Shaanxi University of Science & Technology
关键词
数据挖掘
医学乳腺X线图像
分类
关联规则
神经网络
data mining
digital mammography
classification
association rule
neural networks
