摘要
为乳腺X射线影像计算机辅助诊断做前期预处理,研究了乳腺图像处理中钙化点感兴趣区域提取的问题。在对乳腺X射线图像进行基本的背景分割后,首先运用改进的区域扩张法实现了对乳腺图像中乳腺区域的提取,然后对乳腺区域部分采用改进的反锐化掩模法进行图像增强,突出钙化点区域,再根据含钙化点的特征选取合适的阈值提取出可能含有钙化点的感兴趣区域(ROI)。试验表明,该方法可完成对乳腺影像的ROI提取处理,有助于提高乳腺疾病诊断的准确率。
To preprocess mammograms for computer-aided diagnosis, this paper mainly researches the extraction of calcification region in mammograms. After the background segmentation to mammogram, it firstly makes the pick-up of the mammary region by using the improved area extension, and then uses the method of improved unsharp masking for image enhancement in the pick-up of the mammary region. It highlights the region of calcification points, and picks up the region of interest （ROI） which may contain calcification based on the characteristics of calcification and a suitable threshold. The result shows that this method can complete the extraction of calcification region of interest, helping to improve the accuracy of diagnosis of mammary diseases.
出处
《计算机系统应用》
2010年第2期83-85,66,共4页
Computer Systems & Applications
