上海交通大学学报

上海交通大学学报（自然版）

• 自动化技术、计算机技术 • 上一篇    下一篇

心脏超声虚拟内窥镜图像预处理技术

薛海虹1，陈滨津2，孙锟1,3，余建国2，王威琪2
  

  1. （1. 上海交通大学医学院 附属新华医院儿科心脏中心，上海 200092； 2. 复旦大学 电子工程系，上海 200433；3. 上海交通大学医学院 附属上海儿童医学中心心内科，上海 200127）
  • 收稿日期:2009-11-09 修回日期:1900-01-01 出版日期:2010-08-31 发布日期:2010-08-31

Image Preprocessing of Echocardiographic Virtual Endoscopy System

XUE Haihong1,CHEN Binjin2,SUN Kun1,3,YU Jianguo2,WANG Weiqi2
   

  1. （1. Center of Pediatric Cardiology, Xinhua Hospital, School of Medicine, Shanghai Jiaotong
    University, Shanghai 200092, China; 2. Department of Electronic Engineering, Fudan
    University, Shanghai 200433, China; 3. Department of Pediatric Cardiology, Shanghai
    Children’s Medical Center, School of Medicine, Shanghai Jiaotong University, Shanghai 200127, China）
  • Received:2009-11-09 Revised:1900-01-01 Online:2010-08-31 Published:2010-08-31

摘要： 虚拟内窥镜是近年来兴起的一项崭新医学影像技术，将虚拟现实与科学计算可视化技术相结合，模拟传统的内窥检查过程而生成具有仿真效果的可视化图像.研发的三维心脏超声虚拟内窥镜系统（3DE IESS）可重建三维虚拟环境，使观察者“进入”被检查患者的心脏内部进行检查观察.采用基于同质掩模区域的滤波算法对超声图像斑点噪声进行抑制，应用brFCM算法对降噪后的超声图像进行初始分割，并使用多尺度分析与纹理对比度相结合的分割技术完成最终的图像分割.实验结果表明，在消除斑点噪声的同时，实现对超声图像边缘细节的有效保持，有助于提高三维心脏超声虚拟内窥镜显示效果.

关键词: 虚拟内窥镜, 可视化, 滤波, 图像分割, 三维超声心动图

Abstract: Virtual endoscopy （VE） is a new visualization method using 3D image data sets to simulate the tracks of a conventional endoscope. By using virtual reality computing techniques and visualization in scientific computing techniques, it can assist with the interpretation of 3D data of the heart in space and make it possible to ‘dive’ into the 3D model of the heart. A speckle reduction technique based on the homogeneous mask area filtering algorithm was applied to suppress the speckle noise, and then the denoised image was further processed by using the brFCM method to get an initial segmentation, after that, an integrated technique based on the combination of the multiscale analysis and texturebased contrast was used to obtain the final segmentation. The experimental results show that the proposed method can eliminate the speckle noise well and preserve the edge details of ultrasound images simultaneously and is helpful for visualization of 3DE virtual endoscopy.