Development and validation of an in-silico virtual testing rig for analyzing total knee arthroplasty performance during passive deep flexion: A feasibility study

Abstract

The use of in-silico finite element (FE) models has become more common in orthopedic applications and in the design of biomedical devices, since they can provide results comparable to in vitro experiments while maintaining lower cost. The main downside of this kind of analysis is the high computing time, as it can reach hours or even days to complete; this limitation makes it then not suitable for time-sensitive applications, such as probabilistic analyses or helping clinicians in surgical pre-planning or intra-operative setting. In-silico multibody (MB) simulations, on the other hand, are significantly faster than FE simulations (considering each component of the model as a rigid body); although deformability of each model component is a necessary feature in some applications (e.g. simulation of implant-bone micromotions), several outputs of interest in orthopedic applications, such as implant kinematics and contact forces, do not require a fully deformable model. Therefore, this feasibility study aimed to develop a MB model of a human knee joint implanted with a Total Knee Arthroplasty; a 10 second flexion movement up to 105° was then simulated and the results compared with validated FE models results (under similar boundary conditions) from literature, to perform a preliminary validation in terms of kinematic and kinetic results between the two methods. The agreement and relatively low computing time obtaining with this approach represent a promising starting point for subsequent studies and applications of such techniques in the clinical field.

Keywords: Computing time; FEA; In-silico analysis; Multibody simulations; TKA.