76例乳腺肿瘤超声图像预处理研究

The Research on Preprocessing for the Gray-Scale Ultrasound Breast Tumor Images of 76 Cases

摘要: 以76例乳腺肿瘤灰阶超声图像为研究对象，根据医学超声图像的特点及P-M模型的缺点，提出以图像的局部信息确定扩散门限的改进的P-M模型滤波方法，通过采用多种图像预处理算法及上述改进的P-M模型滤波法对76例乳腺肿瘤超声图像进行试验，实验结果显示，改进的P-M模型滤波方法可以更有效的滤除斑点噪声。

Abstract: This paper mainly focuses on the gray-scale ultrasound breast tumor images. According to the characteristics of ultrasonic image and shortcomings of the P-M model, a modified P-M model filter with local information and spread threshold is proposed. All common pretreatment algorithms are put into experiments and a comparison is made among them. The results show that the modified P-M model filter can more effectively remove the speckle noise.