Virtual Simulation System For Robot Assisted Laryngoscopic Surgery

Virtual surgery simulation is the application of virtual reality technology in modern medicine. Through accurate modeling of surgical instruments and human organs, and realistic simulation of various surgical phenomena such as deformation of organic tissues, surgery simulation system can provide operators with reusable virtual training and simulation environment. Besides, by integrating with surgical robots, virtual surgery simulation could also afford accurate positioning information to master-and-slave motion control of the robot systems.Generally, surgeons of laryngoscopic surgery should possess good ability to coordinate their both hands. In addition, as surgical operations are mainly performed in narrow and long instrumental cavities, appliances of both hands are prone to collide with each other, resulting in operation mistakes, sometimes even dangerous accidents. Thus, surgeons of laryngoscopic surgery need undergo several periods of training. Considering that traditional approaches might not exactly meet demands of surgical training, this dissertation carries out study on surgical simulation system for laryngoscopic surgery, especially on techniques of modeling and collision detection, and successfully builds a virtual simulation system which is based on GHOST SDK consequently.Firstly, this dissertation studies geometric modeling of the virtual simulation system. For virtual surgical instruments, modeling approaches based on 3D CAD software are adapted; for virtual surgical objects, 3D reconstruction methods with respect to their CT image data are studied. On the basis of analyzing mechanical properties of laryngeal organs such as vocal cords, physical model of deformation simulation of vocal cords is then made. Next, this paper performs computational modeling of deformation of vocal cords, constructs two-layer vocal cords using Kelvin–Voigt model, solves status equations of particles system of vocal cords by standard Runge-Kutta approach and gets status of each particle of vocal cords at any time eventually. This dissertation also studies collision detection algorithms based on hierarchical bounding box, analyses characteristics of various collision detection libraries, builds collision detection module through open library ColDet and realizes body-to-body intervention detection as a result. On the basis of theoretical research, this dissertation constructs console of master hands of ENT surgery robot system, which uses PHANToM Desktop to be its master hands, adds program module for periodically sampling poses of master hands, fulfills force feedback constraints of motion paths and makes operators acquire favorable feelings of force feedback.At the end of this paper, experimental study in terms of the developed simulation system for robot assisted laryngoscopic surgery is performed. Experimental results show that this system successfully realizes sampling process of information on poses of master hands of ENT robot system, provides sequent motion control with accurate input data; realistic surgery scene graphs and real-time force feedback constraints of the simulation system also greatly strengthen the safety and stability of surgeons'operations.