基于MRI影像的膀胱肿瘤三维形态变化的初步研究

目的：从膀胱的MRI影像中挖掘膀胱壁的厚度变化信息和形状变化信息，为膀胱肿瘤的计算机辅助诊断提供理论依据。方法：获取膀胱核磁共振图像并进行膀胱壁的手工分割，计算与分析膀胱壁的三维厚度值、相对厚度值及其导数变化，并结合形状指数找出膀胱壁组织与肿瘤组织的形状变化趋势。结果：患者与志愿者之间的膀胱壁的三维厚度值存在统计学差异：当某体素的相对厚度值NBWT〉1．5时，位于肿瘤主体区域的可能性很大：当某体素的相对厚度值的导数为〉0．1或〈一0．1的任意一种情况时，位于肿瘤组织与正常壁组织的过渡区域的可能性很大。利用SI也可达到筛选肿瘤疑似区域的目的。结论：膀胱壁的三维厚度、相对厚度和形状指数可作为区分膀胱壁的正常组织和肿瘤组织的特征指标。

MRI-based 3D Morphological Variation of Bladder Tumor

Objective To provide evidences for diagnoses of bladder tumors by using the information of bladder wall thickness and morphological variation of MRI imaging. Methods MRI images of bladder were collected and the bladder wall was sectioned manually. Then the 3D bladder wall thickness of each voxel was computed to obtain the relative thickness value and correspond- ing first derivative of each voxel and finally the SI value. All the data were analyzed, llesulta There was significant difference be- tween the 3D bladder wall thickness of the patient and that of the volunteer. When the value of NBWT was more than 1.5, the voxel was highly probably located in the main portion of the bladder tumor; when the value of the first derivative of NBWT was more than 0.1 or less than -0.1, the voxel was highly probably located in the transition region. Coachmion The 3D bladder wall thickness, relative thickness and SI can be used as the features to distinguish the abnormal area and normal area of the bladder wall. Chinese Medical Equipment Journal, 2014,35 （3） ： 1-61