Despite decades of clinical research in artificial joints and underlying failure mechanisms, systematical and reproducible identification of reasons for complications in total knee replacements (TKR) remains difficult. Due to the complex dynamic interaction of implant system and biological situs, malfunction eventually leading to failure is multifactorial and remains not fully understood. The aim of present study was to evaluate different TKR designs and positions with regard to joint kinematics and stability under dynamic conditions by using a robot-based hardware-in-the-loop (HiL) setup. An industrial 6-axis robot with 6-axis force-torque sensor mounted into its end-effector moved and loaded real, commercially available TKR (bicondylar, cruciate-retaining) that were in virtual interaction with a subject-specific computational multibody model representing the anatomical situs of the knee joint while performing passive seated deep knee flexion. The subject-specific musculoskeletal multibody model (MMB) included rigid bones of the lower right extremity. Bone and cartilage geometries were reconstructed from MRT/ CT data sets preserving anatomical landmarks and allowing for the calculation of inertial properties. M. quadriceps femoris was modeled as single passive tensile force elements. Knee ligaments were modelled as elastic spring elements with a nonlinear force-displacement characteristic. Providing the flexion angle, the robot moved and loaded the mounted femoral implant component with respect to the tibial component while being in continuous interaction with the MMB. Several influencing parameters like implant position (internal/external rotation, varus/valgus alignment) and design (fixed vs. mobile bearing, tibia-insert height) as well as ligament insufficiency and joint loading on joint kinematics and stability was systematically analysed.Introduction
Material & methods
The purpose of this study was to experimentally evaluate impingement and dislocation of total hip replacements while performing dynamic movements under physiological-like conditions. Therefore, a hardware-in-the-loop setup has been developed, in which a physical hip prosthesis actuated by an industrial robot interacts with an in situ-like environment mimicked by a musculoskeletal multibody simulation-model of the lower extremity. The multibody model of the musculoskeletal system comprised rigid bone segments of the lower right extremity, which were mutually linked by ideal joints, and a trunk. All bone geometries were reconstructed from a computed tomography set preserving anatomical landmarks. Inertia properties were identified based on anthropometric data and by correlating bone density to Hounsfield units. Relevant muscles were modeled as Hill-type elements, passive forces due to capsular tissue have been neglected. Motion data were captured from a healthy subject performing dislocation-associated movements and were fed to the musculoskeletal multibody model. Subsequently, the robot moved and loaded a commercially available total hip prosthesis and closed the loop by feeding the physical contact information back to the simulation model. In this manner, a comprehensive parameter study analyzing the impact of implant position and design, joint loading, soft tissue damage and bone resection was implemented.Introduction
Methods
