摘要
目的利用灰度共生矩阵法提取的纹理特征对新疆高发病哈萨克族食管癌医学图像进行分类研究。方法运用灰度共生矩阵法分析新疆高发病食管癌X钡剂造影医学图像,缩窄型和溃疡型食管癌图像各30张,对图像进行尺寸归一、去噪和增强等预处理,计算0°、45°、90°和135°方向的能量、熵、对比度、相关性和逆差矩的均值和方差,构成特征向量,使用Bayes判别分析法对特征的分类能力进行评价。结果使用Bayes判别分析法对新疆高发病哈萨克族食管癌医学图像进行分类,对缩窄型食管癌图像的分类准确率达到了70%,对溃疡型食管癌图像的分类准确率达到了90%。结论灰度共生矩阵法提取的特征在对不同类型的食管癌图像进行分类时,特征的分类能力有所不同;灰度共生矩阵法可以在一定程度上对不同类型的食管癌进行判别分类。
Objective To discuss the classification on Xinjiang high morbidity of Kazak Esophageal disease based on texture features extracted by GLCM.Methods For X-ray barium angiogram,we normalizing scale,removing the noise by median filter,enhancing by histogram equalization,then we use the gray level co-occurrence matrix (GLCM)to calculate the mean and variance of angular second moment,entro-py,inertia moment,correlation and inverse difference moment in 0°,45 °,90°and 135 ° directions to constitute the texture feature vector,and then evaluate the feature′s classification ability by Bayes discrim-inant analysis.Results Using Bayes discriminant analysis to classify X-ray barium angiogram of Xinjiang high morbidity of Kazak Esophageal disease,the classification accuracy of coarctate esophageal X-ray is 70%,for ulcer esophageal X-ray is 90%.Conclusion The result shows that feature classification ability is different when classifying different images through GLCM which can discriminate the different Esophageal disease,which will help the doctor to diagnosis the Esophageal disease,as well as laying a foundation for computer diagnosis system of Kazak in Xinjiang Uygur autonomous region.
出处
《新疆医科大学学报》
CAS
2014年第3期273-276,共4页
Journal of Xinjiang Medical University
基金
国家自然科学基金(30960097
81160182)
关键词
灰度共生矩阵
新疆
哈萨克族
食管癌
特征提取
图像分类
gray level co-occurrence matrix
Xinjiang high morbidity of Kazak
esophageal disease
feature extraction
image classification