图像分割技术在肝CT容积测量中的应用  

The application of image segment technique in liver volume measurement

作　　者：涂蓉[1] 颜莉蓉[2] 胡德文[2] 伍保忠[1] 黄海伟[1] 

机构地区：[1]海南医学院附属医院放射科,海南海口570102 [2]国防科技大学机电工程与自动化学院,湖南长沙410073

出　　处：《海南医学院学报》2006年第6期498-501,共4页Journal of Hainan Medical University

基　　金：海南省自然科学基金资助项目(30122;30527)

摘　　要：目的:研究一种自动测量肝容积和计算相关指标的计算机技术,并验证其效果。方法:利用基于变形模型的肝轮廓提取方法,在自动识别肝轮廓的基础上,实现自动测量。对15个已知容积的标本进行验证测量,另对12例128幅肝CT图像进行自动分割,同时进行手工和自动测量的肝容积值对照。结果:基于变形模型的图像分割技术可以实现肝容积的自动分割,成功率为68%,在分割成功的图像中,测量的单层肝面积与手工描绘测量的肝面积误差率为5.02%。结论:计算机自动测量技术可以实现肝容积的自动测量,但精确性尚需进一步提高。Objective： To study a method Methods： The algorithm and program of the auto- of extracting liver outline based on modifying modal. distinguished liver outline, the auto-measured liver volume and the auto-calculated related index were explored by using the technique. Results： Liver volume can be automatically measured by tomography, though it requires manual adjustment. The success rate of liver segment was 68%, which was less than one by manual work and in which the error rate of measured liver area was 5.02%. Conclusion： The automated measurement method is feasible, but it needs to be improved on the accuracy and the velocity. They will be a good future in 3D reconstrunction of liver, liver transplantation and so on.

关 键 词：体层摄影术 X线计算机 容积测量 序列图像分割 肝脏 

分 类 号：R735.7[医药卫生—肿瘤] R814.42[医药卫生—临床医学]