Abstract:
The knee is a complex joint, prone to injury and disease. To understand the mechanical behaviour of
the knee and to study its injuries and diseases, computational models can be used. However, the
reproducibility and validity of these models are questionable. The aim was to quantify the influence of
variations in modelling and simulation workflows on the reproducibility of joint level predictions in
computational knee biomechanics. Next to that, the aim was to apply the models to obtain
subject-specific knee ligament moment arms to investigate sexual dimorphism of the knee joint, since
a sex-disparity could contribute to the higher rate of ligament injuries in females compared to males.
Knee joint model development and calibration workflows were established by five modelling teams, as
part of the KneeHub project. A method to compare heterogeneous types of bone and cartilage meshes
was developed and used to compare the meshes between the teams. Next to this, the model
development workflows and the models were compared. After this, we used the model development
and calibration workflows to increase the cohort of models to eight. Then, to study the influence of
model parameters on model outcomes, a sensitivity analysis was performed. Lastly, two male and two
female models were used to obtain the knee ligament moment arms.
First, our workflows were used to build and calibrate two finite element knee joint models. Between
teams, differences in the meshes and mesh generation workflows were found. The differences found
in model development workflows and resulting models showed evidence of the “art of modelling”.
Then, a total of eight models were developed of which four were calibrated adequately. The influence
of the model parameters on model outcomes was model specific and a large sensitivity to ligament
material properties was found. Studying the ligament moment arms, most were smaller in the female
compared to the male models.
Modelling is an iterative and creative process, where a lot of decisions have to be made. Evidence of
the “art of modelling” was found, diminishing modelling reproducibility. Using the models, an indication
of inter-sex differences in ligament moment arms was found.