【摘要】 介绍用共生矩阵和局部灰度差分矩阵对MR图像进行纹理分析,提取熵、粗糙度等纹理特征,对比骨质疏松图像和正常图像,有显著的差别。又对MR图像进行二值化处理,分离骨小梁和骨髓。在此基础上对骨小梁的形态和网络结构进行分析,提取面积、骨架长度和欧拉数等特征参数,对比骨质疏松图像和正常图像,也获得显著的差别。更多还原

【Abstract】 This paper presents a non-invasive and non-ionizing radiation method for assessment of osteoporosis by analyzing in vivo MR images. Texture features (entropy, coarse, etc...) derived from co-occurrence matrix and neighborhood graytone difference matrix are obtained. These features are significantly different between patients and control subjects. Then the grayscale MR image is transformed to the binary image, The shape and topology features (area, skeleton length, euler number, etc.) obtained from the binary image show too significant differences between patients and control subjects.更多还原