基于医学影像处理平台的虚拟膀胱镜系统开发

Development of virtual cystoscopy system based on medical image processing platform

目的:开发一套无创、有效和便捷的虚拟膀胱镜系统,实现膀胱内部的三维显示与观测。方法:基于磁共振成像(MRI)和医学影像处理平台(MITK)类库,建立患者影像信息数据库,并在水平集算法Fast Marching Segmentation及live-wire算法基础上设计图像半自动分割方法,用于实现感兴趣区域提取与三维重建等功能,得到重建后的膀胱腔三维模型,利用模拟内窥镜功能进行多视角观察,实现医生对膀胱腔内病变检测。结果:膀胱腔内三维显示与漫游观测,能够获取肿瘤位置、数量及尺寸等疾病诊断相关的重要信息,实现膀胱肿瘤的术前检测。结论:虚拟膀胱镜的开发,可实现膀胱癌的无创检测与三维可视化观测,为易感人群的临床早期筛查与患者的术后复查提供新的辅助手段。

Objective： To develop a non-invasive, effective, and convenient virtual cystoscopy system so as to realize three-dimensional visualization and observation for the bladder lumen of patient with bladder cancer（BC）. Method： Based on the magnetic resonance imaging（MRI） datasets and medical imaging interaction toolkit（MITK） libraries to establish a database for image information of patients. And based on the arithmetic of Fast Marching Segmentation and arithmetic of live-wire to design semi-automatic partition method of image so as to extract region of interest（ROI）, and realize three-dimension（3 D） reconstruction and other functions, and obtain 3 D mode of bladder lumen pre reconstruction. Using virtual endoscope function to carry out multi-angle observation and realize the detection for lesion of bladder lumen. Results： The 3 D visualization and observation of bladder lumen can obtain importantly diagnosis-related information such as tumor locations, amount and size, and can realize preoperative detection of BC. Conclusion： The development of the virtual cystoscopy system, the non-invasive measurement of visualization BC and 3 D visualization observation can provide new assistant mean for clinically early screening of susceptible population and postoperative reexamination of patients.