Improving accuracy in robotic assisted orthopaedic surgery
Author(s)
Rodriguez y Baena, Ferdinando Maria
Type
Thesis
Abstract
In recent years, medical robotics and computer assisted surgery (CAS) have become the new standard of excellence in many surgical fields, including neurosurgery, minimally invasive heart surgery and general orthopaedics. However, the current state of the art in robotic assistance has yet to reach a stable condition, as researchers across the field are still struggling to meet the constraints associated with the surgical environment. The work described in this thesis focuses on two main areas: the development of novel
solutions to improve the accuracy of robotic assisted surgical orthopaedic procedures, and the design of a robotic system as a concept demonstrator for minimal access Unicompartmental Knee Arthroplasty (UKA). The study builds upon the original implementation of an active-constraint robot (ACROBOT) for Total Knee Arthroplasty (TKA) , described in [Jakopec, 2001]. Early research focuses on the activities which led to the first clinical application of the ACROBOT System, for a number of TKA procedures. An analysis of these early results , which have been published [Jakopec et al., 2001], highlighted the need to tackle accuracy related issues, such as stiffness, bone motion (clamping and motion detection), access and registration (robustness, accuracy, time). These activities, and related results, are presented. The latter aspect, registration , is one of the main topics of this thesis, which has involved the development of an anatomical registration technique based on a point touching scheme, suitable for both open surgery and minimal access knee surgery. Extensive validation through simulation and plastic bone trials has demonstrated that both the algorithm implementation and the robust point acquisition protocol are appropriate for fast and accurate registration of the tibia and femur in both TKA and UKA. The ACROBOT System has been adapted to demonstrate the suitability of a hands-on robot for minimal access knee surgery. Hardware and software modifications are described, while early assessment of the system can be found in [Cobb et al., 2003; Jakopec et al., 2003a] . This thesis also describes the derivation and validation process of a novel hybrid registration algorithm, known as the Bounded ICP algorithm, suitable for minimal access registration of the femur. A discussion of the results and their implications has led to suggestion for future work.
solutions to improve the accuracy of robotic assisted surgical orthopaedic procedures, and the design of a robotic system as a concept demonstrator for minimal access Unicompartmental Knee Arthroplasty (UKA). The study builds upon the original implementation of an active-constraint robot (ACROBOT) for Total Knee Arthroplasty (TKA) , described in [Jakopec, 2001]. Early research focuses on the activities which led to the first clinical application of the ACROBOT System, for a number of TKA procedures. An analysis of these early results , which have been published [Jakopec et al., 2001], highlighted the need to tackle accuracy related issues, such as stiffness, bone motion (clamping and motion detection), access and registration (robustness, accuracy, time). These activities, and related results, are presented. The latter aspect, registration , is one of the main topics of this thesis, which has involved the development of an anatomical registration technique based on a point touching scheme, suitable for both open surgery and minimal access knee surgery. Extensive validation through simulation and plastic bone trials has demonstrated that both the algorithm implementation and the robust point acquisition protocol are appropriate for fast and accurate registration of the tibia and femur in both TKA and UKA. The ACROBOT System has been adapted to demonstrate the suitability of a hands-on robot for minimal access knee surgery. Hardware and software modifications are described, while early assessment of the system can be found in [Cobb et al., 2003; Jakopec et al., 2003a] . This thesis also describes the derivation and validation process of a novel hybrid registration algorithm, known as the Bounded ICP algorithm, suitable for minimal access registration of the femur. A discussion of the results and their implications has led to suggestion for future work.
Version
Open Access
Date Awarded
2004
Advisor
Davies, Professor Brian
Publisher Department
Department of Mechanical Engineering
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)