结合局部最大类间方差和标记筛选的腹部CT肝脏病灶提取

Extraction of Liver Focal Lesions in Abdominal CT Image Based on Local OTSU and Zone Market Screening

目的:针对肝脏图像的特点提取肝脏病灶,为病灶的特征提取做准备。方法:该算法首先在肝脏内部选择包含病灶的局部图像,再采用局部最大类间方差法进行自动阈值预分割,最后采用标记筛选法消除病灶内部空洞和外部小块游离区域。结果:该方法能较好地提取出肝脏病灶。结论:结合局部最大类间方差和标记筛选的方法提取肝脏病灶省力、快速、效果良好。

Objective To prepare for extraction of liver focal lesion in view of its characteristics of images. Methods A region of interest in the liver region comprising the focal lesion to create a local image was selected firstly, then the threshold to pre-segment the local image was got automatically by using local OTSU in clusters algorithm, at last mark sifting methods to get red of inner holes and outer dissociate areas to get the final contour was used. Results The algorithm can efficiently extract the focal liver lesion. Conclusion The way of extraction of liver focal lesions with local OTSU and zone market screening is rapid and effective.