基于人工标记的光学手术导航仪研究

发布时间：2018-03-19 04:24

  本文选题：手术导航仪　切入点：双目立体视觉　出处：《山东大学》2017年硕士论文　论文类型：学位论文

【摘要】：手术导航仪一直存在于人类手术的历史长河之中,并随着人类科学的发展进行着自我变革。21世纪最初的这十多年,信息技术跨越式发展,计算机运算速度几何级增长、医学信息化技术如PET等多维影像技术新起、图像处理和识别技术相关算法日趋成熟、工业技术的跨越性成长,手术的微创化和精密化成为现代医疗技术和临床手术的追求方向,计算机辅助手术逐步走向临床医疗行业研究前沿。计算机辅助手术融合配准术前CT、MRI等医学影像信息和术中对手术工具的空间定位信息,给予手术医生辅助视图以了解手术工具与人体手术区域的相对关系,从而在术前规划中提高前瞻性,在术中增强过程可靠性和精确性。而手术导航仪系统则演变为计算机视觉领域研究的热点之一,由于手术导航仪系统是整个计算机辅助手术系统的中枢模块。出于对当前可见光光学导航仪的改进和优化,本文提出了一种基于ARTag人工标记的导航仪系统实现方案。本文对手术导航仪系统发展历程做了大致概述,结合当下各种流行的手术导航仪系统的实现方式一些缺点和不足,提出了本文的研究内容。本文研究内容主要如下:1、提出了一种基于ARTag人工标记的手术导航仪实现方案。光学手术导航仪系统是当前被认为最具潜力,并被广泛研究的手术导航仪系统方案,其基于双目立体视觉实现手术工具空间定位。本文启发于目前光学导航仪系统复杂、多目标跟踪能力偏弱或者实现冗杂、成本较高等一系列不足。结合增强现实应用中的ARTag人工标记系统,利于双目立体视觉,通过对ARTag人工标记空间跟踪而实现对手术工具跟踪,从而实现手术导航系统功能。2、针对提出的作为标志物的ARTag人工标记,本文提出了一种快速、稳定、精确度高、对光照条件鲁棒性强的标记检测算法。该算法基于边缘检测,通过线段检测、线段融合、线段延伸、标记四边形重构等步骤,实现了高效、快速的标记检测。结合实验,与相似标记系统ARToolKit相关检测算法比较,验证检测算法光照鲁棒性和准确性。3、针对ARTag标记特性,实现标记边角点链获取,并解码标记ID;对手术工具进行标定的前提下,利于双目视觉原理,计算标记空间三维坐标,进一步实现手术工具空间定位跟踪。对手术工具标定方法,我们通过将手术工具绕端点旋转,根据手术工具上标记特征点轨迹在自身球面的原理,拟合计算球心,即工具工作端点三维坐标,进一步结合三维空间到手术工具坐标系的转换矩阵,计算获得端点在手术工具坐标系下坐标,实现手术工具注册。4、针对本文所提出的手术导航仪方案,对该方案系统空间定位稳定性和精确性做出试验并定量分析。系统空间定位稳定性,手术工具工作端点在固定点,手术工具处于不同姿态状况下,计算手术端点跟踪位置的误差来判定跟踪稳定性。系统空间定位精确性,手术工具工作端点在固定已知距离网格点移动,计算手术端点跟踪位置距离误差来判定跟踪精确性。相关系统试验表明本文提出手术导航仪定位系统方案跟踪精度与现有系统精确度在同一数量级,达到系统需求,具有较高的研究和应用价值。
[Abstract]:Surgical navigation system has existed in the operation of human history, and with the development of science of self transformation in.21 century first more than 10 years, the great development of information technology, computer operation speed exponentially, medical information technology such as PET etc. the new imaging technology, image processing and image recognition technology related algorithm is more and more mature and across the growth of industrial technology, and precision of minimally invasive surgery has become the pursuit direction of modern medical technology and clinical surgery, computer assisted surgery to clinical medical industry research forefront. Computer aided surgery registration of preoperative CT, spatial location information of a surgical tool of MRI medical image information and operation, to give the surgeon auxiliary view to understand the relative relationship with human operation tool operation area, so as to improve the forward-looking in preoperative planning, intraoperative in Strengthen the reliability and accuracy of the system. And the surgical navigation will evolve into a research hotspot in the field of computer vision, the surgical navigation system is the central module of the computer aided surgery system. For the improvement and optimization of the visible light optical navigation, this paper proposes an implementation scheme of the navigation system based on ARTag artificial markers in this paper. An overview of the development process to do roughly the surgical navigation system, combined with the implementation of surgical navigation system present a variety of popular some shortcomings and deficiencies, put forward the research content of this paper. The main contents of this paper are as follows: 1, it puts forward a scheme of operation navigator ARTag artificial markers based on optical system is currently considered the surgical navigation system. The most potential, and is widely studied surgical navigation system, which is based on binocular stereo vision to achieve hand Operation tool for spatial positioning. This paper is inspired by the optical navigation system is complex, multi target tracking ability is weak or jumbled, the high cost of a series of problems. Combined with the enhanced ARTag artificial marker system in the practical application, in favor of the binocular stereo vision based on ARTag artificial space tracking and tracking of labeled surgical tools, so as to realize the navigation operation the function of.2 system, puts forward the ARTag as artificial markers markers, this paper presents a fast, stable, high precision, the labeling algorithm robust illumination conditions. The method is based on the edge detection, line detection, line integration, line extending, marking the quadrilateral reconstruction steps to achieve the efficient marker detection fast. According to the experiment, and similar correlation detection algorithm ARToolKit marker system, verify the detection algorithm of illumination robustness and accuracy of.3 for AR. Tag marker characteristics, realize the marking corner point chain acquisition, and decoding the tag ID; calibration of a surgical tool under the premise, to the principle of binocular vision 3D coordinate calculation, mark space, further realize the operation tool for spatial location tracking. The surgical instrument calibration method, we will through the surgical tool around the endpoint rotation, according to the principle of surgical tools mark track feature points in their own sphere, sphere fitting computation, namely tool endpoint 3D coordinates conversion further combining with the 3D coordinates of the surgical tool to calculate matrix, obtain the endpoint coordinates in the surgical tool coordinate system, the realization of the operation tool for surgical navigation registration.4, the proposed scheme, make quantitative analysis test and the scheme of the system is stable and accurate. The space positioning system of space positioning stability, operation tools work at the end of the fixed point operation tool 浜庝笉鍚屽Э鎬佺姸鍐典笅,璁＄畻鎵嬫湳绔�鐐硅窡韪�浣嶇疆鐨勮��宸�鏉ュ垽瀹氳窡韪�绋冲畾鎬，

本文编号：1632830