摘要
该文根据乳腺肿瘤超声图像在灰度空间的分布差异性,采用共生矩阵对乳腺癌及乳腺增生超声图像进行纹理特征提取,并采用统计学方法进行分析,结果发现乳腺癌和乳腺增生超声图像在能量和对比度上存在显著性差异,具有统计学意义,将有助于进一步对乳腺肿瘤超声图像进行分类和检索。
In this paper,based on differences of breast tumor ultrasound image grayscale distribution,using the gray level co-oc currence matrix of ultrasound images to extract texture feature of breast cancer and breast hyperplasia,and the use of statistical methods for significance analysis.The results show that ultrasound images of breast cancer and breast hyperplasia in energy and contrast,there was significant difference statistically significant,which help to further classify and retrieve ultrasound images.
出处
《电脑知识与技术》
2012年第11期7552-7554,共3页
Computer Knowledge and Technology
基金
新疆医科大学创新基金（XJC201089）
关键词
乳腺超声图像
灰度共生矩阵
特征分析
breast ultrasound image
gray level co-occurrence matrix
feature analysis
