Effect of implementing magnetic resonance imaging for patient-specific OpenSim models on lower-body kinematics and knee ligament lengths. (23rd January 2019)

• Record Type: Journal Article
• Authors: Smale, Kenneth B.; Conconi, Michele; Sancisi, Nicola; Krogsgaard, Michael; Alkjaer, Tine; Parenti-Castelli, Vincenzo; Benoit, Daniel L.
• Is Part Of: Journal of biomechanics. Volume 83(2019)
• Page Start: 9
• Page End: 15
• Publication Date: 2019-01-23
• Subjects: Biomechanics -- Computer simulation -- Knee injuries -- Musculoskeletal model; Animal mechanics -- Periodicals; Biomechanics -- Periodicals; Biomechanics -- Periodicals; Mécanique animale -- Périodiques; Biomécanique -- Périodiques; Electronic journals; 571.4305
• DOI: 10.1016/j.jbiomech.2018.11.016
• Languages: English
• ISSNs: 0021-9290

Abstract

Abstract: Background: OpenSim models are typically based on cadaver findings that are generalized to represent a wide range of populations, which curbs their validity. Patient-specific modelling through incorporating magnetic resonance imaging (MRI) improves the model's biofidelity with respect to joint alignment and articulations, muscle wrapping, and ligament insertions. The purpose of this study was to determine if the inclusion of an MRI-based knee model would elicit differences in lower limb kinematics and resulting knee ligament lengths during a side cut task. Methods: Eleven participants were analyzed with the popular Rajagopal OpenSim model, two variations of the same model to include three and six degrees of freedom knee (DOF), and a fourth version featuring a four DOF MRI-based knee model. These four models were used in an inverse kinematics analysis of a side cut task and the resulting lower limb kinematics and knee ligament lengths were analyzed. Results: The MRI-based model was more responsive to the movement task than the original Rajagopal model while less susceptible to soft tissue artifact than the unconstrained six DOF model. Ligament isometry was greatest in the original Rajagopal model and smallest in the six DOF model. Conclusions: When using musculoskeletal modelling software, one must acutely consider the model choice as the resulting kinematics and ligament lengths are dependent on this decision. The MRI-based knee model is responsive to the kinematics and ligament lengths of highly dynamic tasks and may prove to be the most valid option for continuing with late-stage modelling operations such as static optimization.