微创手术机器人图像系统设计方法与空间映射策略研究
发布时间:2018-06-13 12:50
本文选题:微创手术机器人 + 图像系统 ; 参考:《天津大学》2016年博士论文
【摘要】:图像系统是微创手术机器人获得患者实时信息反馈的重要部分,其优劣不仅决定了微创手术机器人的整体性能,也对机器人其他组成模块的设计与开发有重要影响。微创手术机器人图像系统研究包括内窥镜设计,医生的观察模式与操作模式,主从空间映射等诸多问题;由此,本文开展面向微创手术机器人图像系统设计实现及映射策略研究。主要取得以下创造性成果:1.根据对人类深度知觉的分析,创造性地提出将运动视差概念引入微创手术机器人图像系统。利用运动视差补偿了双目视觉深度信息随视点位移产生的畸变,深入探讨了影响运动视差映射系数的因素,并设计实验测定了适用于微创手术机器人的运动视差映射系数。2.通过对内窥镜探查空间与术野空间的映射关系,提出全维视角立体内窥镜概念。根据微创手术空间/安全性约束,提出一种适用于微创手术机器人内窥镜的柔性关节。将全维视角立体内窥镜原理应用于微创手术机器人“Microhand S”系统,建立了相应的控制算法。3.为提高机器人操作性,提出了将操作空间与视觉空间相融合的映射方法。基于对手部动作的分解,提出一种适用于上述映射方法的非机械结构主操作手模式。建立了符合人体直觉的映射模型,并实现了基于该模型的微创手术机器人主从映射算法和控制流程。4.研究了医学数字影像的空间映射算法。针对医学超声影像,构建了二维与三维空间映射模型,并编写了软件对模型进行验证。建立了超声影像与内窥镜图像的映射算法,实现了多维医学影像的融合。结合医疗手术机器人的特点,分析了适用于机器人的医学影像标定方式。5.在实验研究方面,对全维视角立体内窥镜的可靠性进行了验证,利用模型实验对运动视差在微创手术机器人操作中的作用进行了测定和验证,并通过比对实验对超声影像映射算法进行了测试,通过动物实验验证了多维医学影像融合算法的有效性。
[Abstract]:Image system is an important part for minimally invasive surgical robot to obtain real-time information feedback from patients. Its merits and demerits not only determine the overall performance of the minimally invasive surgery robot, but also have an important impact on the design and development of other components of the robot. The research on image system of minimally invasive surgery robot includes endoscope design, doctor's observation mode and operation mode, master-slave space mapping and so on. In this paper, the design and implementation of image system and mapping strategy for minimally invasive surgical robot are studied. The main achievements are as follows: 1: 1. According to the analysis of human depth perception, the concept of motion parallax is creatively introduced into the image system of minimally invasive surgical robot. Motion parallax is used to compensate the distortion of binocular visual depth information caused by viewpoint displacement, and the factors that affect the mapping coefficient of motion parallax are discussed in depth. The kinematic parallax mapping coefficient of minimally invasive manipulators is measured by design experiments. Based on the mapping between the exploration space of endoscope and the space of surgical field, the concept of full-dimensional perspective stereo endoscope is put forward. According to the space / safety constraints of minimally invasive surgery, a flexible joint suitable for endoscope of minimally invasive surgery robot is proposed. The principle of full dimensional stereoscopic endoscope is applied to the microhand S system of minimally invasive surgery robot, and the corresponding control algorithm .3. In order to improve the maneuverability of the robot, a mapping method is proposed, which combines the operation space with the visual space. Based on the decomposition of hand movements, a non-mechanical structure master manipulator model is proposed for the above mapping methods. The mapping model of human body intuition is established, and the master-slave mapping algorithm and control flow of minimally invasive surgical robot based on this model are implemented. The spatial mapping algorithm of medical digital image is studied. A two-dimensional and three-dimensional mapping model was constructed for medical ultrasound images and a software was developed to verify the model. The mapping algorithm of ultrasound image and endoscope image is established, and the fusion of multidimensional medical image is realized. Combined with the characteristics of medical surgical robot, the medical image calibration method suitable for robot is analyzed. In the aspect of experimental research, the reliability of full dimensional stereoscopic endoscope is verified, and the function of motion parallax in the operation of minimally invasive surgical robot is measured and verified by model experiment. The ultrasonic image mapping algorithm is tested by comparison experiment, and the validity of multi-dimensional medical image fusion algorithm is verified by animal experiments.
【学位授予单位】:天津大学
【学位级别】:博士
【学位授予年份】:2016
【分类号】:TP391.41;TP242
【参考文献】
相关期刊论文 前3条
1 巫影,何琳,黄映云,朱海潮,吕志强;虚拟现实技术综述[J];计算机与数字工程;2002年03期
2 吕平,刘芳,戚昭恩;腹腔镜外科百年发展史[J];中华医史杂志;2001年04期
3 田增民,王田苗,刘宗惠,杜吉祥,赵全军,陈卫东;机器人系统辅助脑立体定向手术[J];军医进修学院学报;1998年01期
,本文编号:2014082
本文链接:https://www.wllwen.com/shoufeilunwen/xxkjbs/2014082.html