Development of a knee simulator for ligament balancing

Introduction. The knee joint is a result of a complex interaction of tissues, including bones, muscles, tendons, ligaments and cartilage. In a pathological situation the function of the knee has to be reestablished using a knee prothesis. The goals of total knee arthroplasty (TKA) are to eliminate pain and re-establish a stable and balanced knee. From 2001 to 2016, 812 639 primary TKA were performed in Italy. Although most cases are successful, patients complain about instable knee joint during daily activities. A key element concerns the balancing of the knee, in particular the ligament balancing, a procedure affecting mainly medial (MCL) and lateral (LCL) collateral ligaments. The objective of the thesis is to develop a knee simulator to simulate different ligament tensions and allow the measurement of quantitative parameters. Materials and methods. The simulator is an aluminum skeleton that reproduce a lower limb for fixing plastic models of tibia and femur. A system for measuring ligament tension was designed using a microprocessor and two load cells, one for each collateral ligament. Then, a measurement system was used to evaluate the intercompartmental forces between the joint surfaces. The design of the simulator also concerns the search for a synthetic material that reproduces the mechanical properties of the physiological ligaments: mechanical properties of elastomeric samples produced by 3D printing were tested using uniaxial tensile tests. Results. First, in-depth bibliographic research generates reference values regarding the mechanical properties and normal functioning conditions of physiological ligaments. Ligament stiffness were around 60-70 N/mm, ligament tension around 50 N considering both ligaments (higher values are considered during surgery), intercompartmental forces around 20-180 N. Both our ligament tension and intercompartmental forces measurement systems produce results within the reference ranges. Data were recorded at 0 and 90 degrees of flexion. With a ligament tension around 10 N at 90° of flexion, a tension of 20-25 N was reached in full extension. With these values, intercompartmental forces of 25-40 N at 0° of flexion and 90-120 N in full extension were evaluated. As expected, maximum values of ligament tension and forces were recorded at 0° of flexion. Linear springs were used as synthetic ligaments because results of uniaxial tensile tests on surrogate ligaments show viscoelastic behavior but inferior mechanical properties: maximum force was less than 12 N instead of hundreds of newtons. Discussion. A first version of a simulator was successfully designed and developed from the early stages. It allows the fixing of plastic bone models, variation of leg flexion angle and simulation of different varus/valgus angles. The ligament tension measurement system uses commercially available low-cost components, entirely assembled in our laboratory. All measurements are acquired using linear springs instead of synthetic ligaments because additional tests and analysis must be conducted on elastomers to improve prototyping of a material suitable for replicating the mechanical properties of physiological ligaments. Both ligament tension and intercompartmental forces measurement system proved to be within reference range from literature. The simulator needs to be further refined in the next version, in terms of measurements precision of the two load cells, and stepper motors may be inserted to tension the ligaments in an automatic and more reproducible way.