中国组织工程研究 ›› 2011, Vol. 15 ›› Issue (4): 745-748.doi: 10.3969/j.issn.1673-8225.2011.04.043

• 骨与关节综述 bone and joint review • 上一篇    下一篇

计算机辅助医学图像三维重建的算法分析

王瑞娟1,张  季1,彭  可2   

  1. 1南京中医药大学信息技术学院,江苏省南京市 210046
    2东华软件股份公司,北京市100190
  • 收稿日期:2010-10-20 修回日期:2010-12-13 出版日期:2011-01-22 发布日期:2011-01-22
  • 作者简介:王瑞娟★,女,1980年生,江苏省常州市人,汉族,2010年南京理工大学毕业,硕士,讲师,主要从事计算机网络、医学图像处理的研究。 wangruijuan_2005@126.com

Algorithm for computer-assisted medical image 3D reconstruction

Wang Rui-juan1, Zhang Ji1, Peng Ke2   

  1. 1Information Technology Institute of Nanjing University of Traditional Chinese Medicine, Nanjing  210046, Jiangsu Province, China
    2DHC Software Co., Ltd., Beijing 100190, China
  • Received:2010-10-20 Revised:2010-12-13 Online:2011-01-22 Published:2011-01-22
  • About author:Wang Rui-juan★, Master, Lecturer, Information Technology Institute of Nanjing University of Traditional Chinese Medicine, Nanjing 210046, Jiangsu Province, China wangruijuan_2005@126.com

摘要:

背景:三维重建技术是采用计算机技术对二维医学图像进行边界识别,重新还原出被检组织或器官的三维图像。
目的:分析在不同情况下进行医学图像三维重建时如何进行算法的选择。。
方法:采用计算机检索中国期刊全文数据库和Pubmed 数据库。中文检索词为“医学图像,三维重建,面绘制,体绘制”,英文检索词为“medical images, three-dimensional reconstruction, surface rendering, volume rendering”。 检索与医学图像三维重建算法相关的文献33篇,从面绘制重置方法和体绘制重置方法的实现原理、实现复杂度、实时显示情况等方面进行分析。
结果与结论:目前,医学图像三维重建根据绘制过程中数据描述方法的不同可分为三大类:面绘制方法、体绘制方法和混合绘制方法。通过对面绘制和体绘制方法中不同算法的分析,可以看到面绘制方法在算法效率和实时交互性上是优于体绘制的,虽然面绘制方法在绘制时候会丢失许多细节,使得绘制图像效果不理想,但是由于其算法比较简单,占用内存资源少,所以目前得到了广泛的运用。体绘制方法是对体数据场中的体素进行直接操作,可以绘制出三维数据场中更丰富的信息,因此体绘制方法的绘制效果优于面绘制方法。

关键词: 医学图像, 三维重建, 可视化, 面绘制, 体绘制

Abstract:

BACKGROUND: Three-dimensional reconstruction identifies two-dimensional medical images to recover the three-dimensional images of detected tissues and organs.
OBJECTIVE: To analyze algorithm choice for medical image 3D reconstruction.
METHODS: A computer-based online search of CNKI and Pubmed databases was performed to search related articles with key words “medical image, three-dimensional reconstruction, surface rendering, volume rendering” in Chinese and English. A total of 33 articles regarding algorithm of medical image three-dimensional reconstruction were collected to analyze the implementation principle, complexity, and real-time display of surface rendering technology and volume rendering technology.
RESULTS AND CONCLUSION: Medical image three-dimensional reconstruction is classified into surface rendering, volume rendering and combined rendering according to data description methods. Surface rendering exhibits advantages in algorithm efficiency and real time reciprocity compared with volume rendering. Although surface rendering results in detail loss during rendering, it has been extensively applied due to simple operation and small memory occupying. Volume rendering is direct operation on voxel in volume dataset, which can render abundant information of three-dimensional dataset. Therefore, the rendering effect is better than surface rendering.

中图分类号: