Understanding how structural and functional alterations of individual tissues impact on whole-joint function is challenging, particularly in humans where direct invasive experimentation is difficult. Finite element computational models produce quantitative predictions of the mechanical and physiological behaviour of multiple tissues simultaneously, thereby providing a means to study changes that occur through healthy ageing and disease such as osteoarthritis. As a result significant research investment has been placed in developing such models of the human knee. Previous work has highlighted that model predictions are highly sensitive to the various inputs used to build them, particularly the mathematical definition of material properties of biological tissues. The goal of this systematic review is two-fold. First, we provide a comprehensive summation and evaluation of existing material property data for human knee joint tissues, tabulating numerical values as a reference resource for future studies. Second, we review efforts to model whole-knee joint mechanical behaviour through finite element modelling with particular focus on how studies have sourced tissue material properties. The last decade has seen a renaissance in material testing fueled by development of a variety of new engineering techniques that allow the mechanical behaviour of both soft and hard tissues to be characterised at a spectrum of scales from nano- to bulk tissue level. As a result there now exists an extremely broad range of published values for human knee tissues. However, our systematic review highlights gaps and ambiguities that mean quantitative understanding of how tissue material properties alter with age and osteoarthritis is limited. It is therefore currently challenging to construct finite element models of the knee that are truly representative of a specific age or disease-state. Consequently, recent whole-joint finite element models have been highly generic in terms of material properties even relying on non-human data from multiple species. We highlight this by critically evaluating current ability to quantitatively compare and model 1) young and old and 2) healthy and osteoarthritis human knee joints. We suggest that future research into both healthy and diseased knee function will benefit greatly from a subject- or cohort-specific approach in which finite element models are constructed using material properties, medical imagery and loading data from cohorts with consistent demographics and/or disease states.
The peripheral soft tissues should not be ignored in the finite element models of the human knee joint
