基于多期MR影像的肾脏自动定位与分割方法  

Effective Automated Kidney Locating and Segmentation Method Based on Multi-phase MR Images

作　　者：李宏[1] 潘晓光[2] 孙浩然[3] 康雁[1] 

机构地区：[1]东北大学中荷生物医学与信息工程学院,辽宁沈阳110819 [2]东北大学信息科学与工程学院,辽宁沈阳110819 [3]天津医科大学总医院放射科,天津300052

出　　处：《东北大学学报（自然科学版）》2013年第1期52-56,共5页Journal of Northeastern University(Natural Science)

基　　金：国家自然科学基金资助项目(61071213;81171316);天津市自然科学基金资助项目(08JCYBJC07900)

摘　　要：肾脏自动定位和分割是肾脏功能分析流程中的重要一步,针对此提出了一种基于多期MR影像的肾脏自动定位及分割方法,即利用肾脏在多期MR扫描影像中表现出的特性,通过三维配准减影及最大密度投影方法,得到数据体在x,y,z方向的投影图像,并利用最大密度分布曲线的特性,得到肾脏区域的三维定位;在定位区域内利用水平集方法对实质期影像进行肾脏提取,然后将分割结果映射到其余多期影像上.实验结果表明,该方法得到的肾脏定位结果准确,且在定位区域内完成的肾脏分割执行效率高,分割结果较好.Automatically locating the kidney region and segmenting the kidneys in MR images are the first and very important steps for the whole renal function analysis flow.A novel method of automatically locating the kidney based on multi-phase MR images was proposed.First,3D registration and subtraction were employed;then max intensity projection along different directions was performed to get projection images;afterwards,the MID（maximum intensity distribution） curve analysis was used to obtain the kidney region boundary.Inside the kidney region,the level set method was employed to segment each kidney on the parenchymal phase images.Finally,the segmentation result was mapped onto all the other phase images.Experimental results showed that the location results are very good,and the segmentation performed within the location region has higher efficiency and better accuracy.

关 键 词：磁共振影像 肾脏分割 最大密度分布曲线 水平集 肾小球滤过率 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]