基于肺部CT的肿瘤形态与纹理的三维变化检测  

3DChange Detection of Pulmonary Tumor in Shape and Texture Based on Lung CT

作　　者：熊婷[1] 黄绍辉[1] 黄晓阳[1] 王连生[1] 王博亮[1] 

机构地区：[1]厦门大学信息科学与技术学院,福建厦门361005

出　　处：《厦门大学学报（自然科学版）》2016年第3期401-405,共5页Journal of Xiamen University：Natural Science

基　　金：国家自然科学基金(61001144;61102137;61301010;61327001)

摘　　要：提取4个阶段肺部肿瘤CT图像的肿瘤区域,进行三维重建,计算三维肿瘤的形态特征参数和纹理特征参数,研究这两部分参数与肿瘤的变化关系.基于体素和Quick Hull算法计算出肿瘤的形态特征参数,结合灰度共生矩阵计算出4个阶段三维肿瘤的纹理特征参数,组成判别肿瘤三维变化的特征参数组合.实验结果证明,形态特征参数可以判断肿瘤的体积大小变化,纹理特征参数可以判断肿瘤内部活性变化.肺部肿瘤在治疗阶段体积大小发生了先变小后变大的变化,肿瘤纹理均匀性和粗细程度也发生了变化,在阶段Ⅲ体积和最长径的值最小,共生矩阵的能量、熵值和逆差分矩最大,这说明肿瘤最小且内部发生了坏死,这与临床结论相一致,为进一步指导肿瘤疗效评估奠定了基础.It is a challenging task to evaluate the tumor treatment effect from CT images, because the task mainly depends on experi- ences of doctors.For the purpose of establishing an objective evaluation method, a 3D tumor change detection method based on shapes and texturesis proposed in this paper.First, we reconstruct the 3D tumor in four periods using VTK, then calculate 3D tumor＇s char- acteristic parameters,which are used to compare with the change of cancer during four periods.The experimental result shows that shape-characteristic parameters based on voxel and Quick Hull can be used to judge the size change of tumor,and that texture-characteristic parameters based on gray level co-occurrence matrix can be used to compare with the change of tumor internal activity.In period Ⅲ , the volume and longest diameter of tumor reaches the minimum;The energy, entropy and inverse-difference moment reach the maximum.This phenomenon explains that the tumor size is the least and the tumor becomes liquefied.This conclusion is consist- ent with the clinical validation.

关 键 词：肺部肿瘤 CT图像 Quick HULL 纹理特征 灰度共生矩阵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]