Tibiofemoral Model

The tibiofemoral model that is used in the optimization scheme is based on the model that was developed in the Model Development phase of this project. This section documents any changes that were made to the original model to facilitate the calibration procedure.

Given the model will likely be evaluated hundreds of times, specific anatomy were removed to increase computational efficiency. The patella is unloaded during the experimental laxity tests. As a result, the patella and the patellar ligaments were removed from the knee model. Additionally, the medial and lateral meniscus and mensical attachments were also removed. It is recognized that removing the meniscus may affect the estimate ligament slack lengths due to its potential to provide joint restraint. For the relatively low magnitudes of loading utilized in this work, it is assumed that minimal restraint is provided by the meniscus. This assumption will be tested during confirmation of the calibration procedure and updated if necessary.

The following sections offer more detail on the model that is used in the optimization scheme, though do note that an error was discovered since the Model Development package was disseminated (outlined below). Also note that unlike the Open Knees data, all registration was performed during the Model Development phase and has not been reproduced here.

Ligament Properties

Each ligament bundle’s material properties include two slack length values, toe region percent, and stiffness. The calibration procedure will target bundle-specific slack lengths. The toe region percent and the stiffness values are not included in the calibration procedure and will be set to values defined during Model Development. In short, ligament stiffness values are defined based on literature values and all toe regions are be set at 6% strain.

Model Development Deviation

A mistake was found in the plACL insertion (specifically the file dat/DU02/mriData/Segmentation/plAclLine.fcsv which is located in the compressed directory titled DU02ModelDevelopment_IntermediateOutputs.tar.gz). The mistake likely occurred in 3D Slicer where the insertion point may have been moved when rotating the model in the 3D viewing window. This mistake is reflected in the Model Development package.

Fortunately the file is in a local repository, and the original version of the file was restored. Fig. 1 shows a comparison of plACL insertion point that was mistakenly moved and reported in the Model Development package, and the restored insertion point that is used in the model for calibration.

control variable

Fig. 1 A medial view of the right knee model showing (yellow) the plACL insertion point that was mistakenly moved, and is reported in the Model Development intermediate outputs, and is reflected in the plACL mesh. (red) The plACL insertion point before it was mistakenly moved. The red point shows the location of the plACL insertion point that is used in model calibration.