Novel Strategies for Enhanced Navigation in Robotic Surgical System

手術導航技術為外科醫生提供了相對於醫療影像的手術工具及病灶的準確的位姿信息，可以擺脫在傳統手術中對於醫生經驗與視野的高度依賴。因此，採用導航系統進行手術可以極大地提高手術精準度及效率。跟蹤定位系統是手術導航系統的靈魂。在廣泛地評價現有及正處於製階段的跟蹤定位系統之後，通過比較並分析了他們的利弊我們將這些技術中存在的問題及挑戰總結為信息缺失問題。本論文著眼于手術導航系統的信息缺失問題，特別是在當前被廣泛使用的光學跟蹤定位系統(OTS) 的視線遮擋(LoS) 以及視野限制(FoV)問題。在研究背景一章中，我們考察了信息缺失問題的現有解決方案並總結了他們的特性及限制因素。對於導航系統在解決信息缺失方面的發展前景，我們從新型傳感技術、智能及感知、機器人輔助的跟蹤定位適應性策略等方面進行了討論。為解決OTS系統的視線遮擋及視野限制問題，我們提出了一種基於多個可重新配置的單目模塊的跟蹤導航系統。我們首先介紹了系統架構及一種適用於手術工具的快速多相機標定方法。之後，我們採取數據融合方法，使用多個模塊對手術工具進行定位跟蹤。通過對多模塊方法的研究，我們發現單純利用靜態的導航方法無法完全解決光學系統中的信息缺失問題。為此，我們研製了一套機器人跟蹤定位系統，該系統在遮擋物阻擋視線及手術工具離開視野時可以自主操控光學跟蹤設備。我們搭建了一套由一個4自由度PPRR機器人、一個雙目光學跟蹤設備和一個RGB-D相機組成的機器人系統平台，並提出了一系列基於最優化的控制方法為光學跟蹤提供了一個流暢的手術過程體驗及適應性視野。通過大量仿真及實驗，我們給出了在真實手術環境中的不同中斷場景下的量化分析來驗證該系統的可行性。為進一步提高機器人跟蹤系統的性能，我們還研究了一種協作式的機器人跟蹤方法。我們提出了一套由四個3 自由度PRR機器人組成的系統，每個機器人上都安裝有一個光學跟蹤設備。在四個機器人底座的中心位置放置一個RGB-D 相機來對手術場景進行分析。因為該系統具有協作性，為了提高系統效率及最小化累計誤差，對四個機器人空間位置的標定成為了不可或缺的一部分。我們採用配對方式將該標定問題表述為一個AXB = YCZ 問題，並提出了一種同步標定方法。綜上所述，本文研究了手術導航系統中使用的光學定位技術中最重要及最富挑戰性的信息缺失問題。提出了若干解決此問題及附屬子問題的新型方案。大量仿真及真實實驗的結果驗證了我們提出的方案及算法有效性及可行性。Surgical navigation technology provides surgeons with the relative pose relationship amongst medical images, surgical instruments and lesions, compensates the drawbacks of the traditional operation which depends heavily on the surgeon experience and sight. Consequently, the introduction of navigation system will lead to accurate, precise and efficient operations. Tracking system is the soul of surgical navigation system. After reviewing the tracking technologies that are currently available or under research, we compare and analyze their pros and cons, and find a common challenge for these technologies, i.e. the information loss problem. This thesis focuses on the information loss problem in the navigation systems, especially the line-of-sight (LoS) occlusion problem and the limited field-of-view (FoV) in the optical tracking system (OTS) which has been regarded as thede-facto standard in current operative rooms.As a background study, we survey the existing solutions of the information loss problem, and summarize their key properties and limitations. The possible research trends for the navigation systems to solve the information loss problem, including novel sensing technology, intelligence and perception capability, and adaptive strategy for robot-assisted tracking are discussed.To solve the occlusion and limited FoV problems of OTS, we propose a tracking system based on multiple reconfigurable monocular modules. Novel multi-modular system architecture and thecorresponding agile multi-module calibration method which only uses the surgical instrument are introduced. A robust data fusion algorithm is also proposed to enable the surgical instrument tracking by multiple modules. Extensive simulations and experiments have been carried out and the superiority of the proposed calibration and fusion algorithms has been illustrated. A cadaver experiment has been carried out and the results have validated the feasibility of the proposed system, calibration algorithm and the robust data fusion algorithm in a real-world surgical scenario.After learning the limitations from the multi-modular approach, we investigate a robotic tracking system that autonomously maneuvers the optical tracker when obstacles are blocking the LoS or instrument moving out of the FoV. We propose a robotic platTo further improve the performance of the robotic tracking system, a cooperative robotic approach is also investigated. We propose a robotic platform consisted of four 3DoF PRR robots, each of which is mounted with an optical tracker. A RGB-D camera iIn conclusion, this thesis investigates the most important and challenging information loss problem of the optical tracking technology for the surgical navigation system used in a robotic surgery. Several novel schemes to solve the problem and the corWang, Jiaole.Thesis Ph.D. Chinese University of Hong Kong 2016.Includes bibliographical references (leaves ).Abstracts also in Chinese.Title from PDF title page (viewed on …).Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only