计算机应用 ›› 2021, Vol. 41 ›› Issue (3): 833-838.DOI: 10.11772/j.issn.1001-9081.2020060776

所属专题: 多媒体计算与计算机仿真

• 多媒体计算与计算机仿真 • 上一篇    下一篇

基于增强现实的手术导航系统快速标定算法

孙启昌, 麦永锋, 陈晓军   

  1. 上海交通大学 机械与动力工程学院, 上海 200240
  • 收稿日期:2020-06-09 修回日期:2020-10-13 出版日期:2021-03-10 发布日期:2020-12-22
  • 通讯作者: 陈晓军
  • 作者简介:孙启昌(1995-),男,山东枣庄人,硕士研究生,主要研究方向:增强现实、计算机辅助手术;麦永锋(1995-),男,广东佛山人,硕士研究生,主要研究方向:计算机辅助手术、图像配准;陈晓军(1976-),男,湖北黄冈人,研究员,博士,主要研究方向:手术导航、医学图像处理。
  • 基金资助:
    国家重点研发计划项目(2017YFB1302903);国家自然科学基金资助项目(81971709,M-0019,81828003,82011530141);上海市科委“科技创新行动计划”(19510712200,20490740700);上海交通大学医工交叉基金资助项目(ZH20182DA15,YG2019ZDA06,ZH2018QNA23);厦门市重大科技项目(3502Z20201030)。

Fast calibration algorithm in surgical navigation system based on augmented reality

SUN Qichang, MAI Yongfeng, CHEN Xiaojun   

  1. School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
  • Received:2020-06-09 Revised:2020-10-13 Online:2021-03-10 Published:2020-12-22
  • Supported by:
    This work is partially supported by the National Key Research and Development Program of China (2017YFB1302903), the National Natural Science Foundation of China (81971709, M-0019,81828003, 82011530141), the Foundation of Science and Technology Commission of Shanghai Municipality (19510712200, 20490740700), the Shanghai Jiao Tong University Foundation on Medical and Technological Joint Science Research (ZH20182DA15, YG2019ZDA06, ZH2018QNA23), the Key Research Project of Xiamen Municipal Government (3502Z20201030).

摘要: 针对光学半透式头戴显示器(OST-HMD)在增强现实(AR)中的虚实融合问题,提出了一种基于光学定位跟踪系统的OST-HMD快速标定算法。首先,通过把OST-HMD中的虚拟标记点和其在真实世界中的对应点组成两个三维点集,对虚拟空间到光学定位追踪空间之间的映射进行估计,从而求解虚拟空间到真实场景的转换矩阵。然后,建立整个导航系统的矩阵传递关系,在此基础上设计并实现了基于AR的手术导航系统,并对其进行了精度验证实验和模型实验。实验结果表明,所提算法使虚拟基准点和对应的真实基准点之间的均方根误差达到1.39 ±0.49 mm,标定过程的平均耗时为23.8 s,证明该算法具有临床应用的潜力。

关键词: 增强现实, 图像配准, 标定, HoloLens, 手术导航

Abstract: To solve the problem of fusion of virtual scene and real scene of Optical See-Through Head-Mounted Display (OST-HMD) in Augmented Reality (AR), a fast calibration method for OST-HMD was proposed on the basis of optical positioning and tracking system. Firstly, the virtual makers in OST-HMD and the corresponding points in real world were collected as two 3D point datasets, and the transformation between the virtual space and optical positioning and tracking space was estimated to solve the transformation matrix from the virtual space to the real scene. Then, the transitive relation among matrices of the entire navigation system was built, and an AR-based surgical navigation system was designed and implemented on this basis, and the accuracy validation experiment and model experiment were conducted to this system. Experimental results show that the proposed algorithm makes the root mean square error between the virtual datum points and the corresponding real datum points achieved 1.39 ±0.49 mm, and the average time of calibration process of 23.8 seconds, which demonstrates the algorithm has the potential in clinical applications.

Key words: Augmented Reality (AR), image registration, calibration, HoloLens, surgical navigation

中图分类号: