Lanterna Rally

Author: Mousa Kazemi

Publisher:

Published: 2018

Total Pages: 201

ISBN-13:

This work presents a foundation to combine computational musculoskeletal modelling and finite element modelling to investigate joint biomechanics with application to knee osteoarthritis. A novel framework to rapidly generate a finite element model of the knee is developed that comprises subject-specific anatomy, combining data from medical images and motion capture. This work demonstrates the robustness and feasibility of the workflow to predict the three-dimensional kinematics of the joint from the resulting anatomical model, thus addressing the form-function relationship inherent in joint mechanics. Knee joint mechanics are influenced by the geometry of the knee, and in particular the articulating surfaces of the tibial, patellar, and femoral cartilage. The workflow developed in Chapter 3 reconstructs major lower limb bone segments (pelvis, femur, tibia, fibula, and patella), knee cartilage volumes and major ligamentous constraints of the knee, using only sparse motion capture markers and a 12cm field of view MRI scan of the knee. The workflow reconstructed the subject-specific knee finite element model within 25 minutes, which is considered an acceptable time frame for clinical use. The accuracy of this new method is comparable to that achieved by a more traditional approach of generating meshes directly from segmented MR images. Chapter 4 explains principal component analysis and partial least squares regression techniques used to generate statistical shape models of the knee joint. The subject-specific 3D knee models were reconstructed from easily measured anthropometric and demographic measurements. In Chapter 5, a robust tool to reproduce the knee joint geometry from sparse weight-bearing MR imaging data was developed and used to estimate the tibiofemoral and patellofemoral kinematics. The results suggested that the knee joint kinematics may differ for men and women. The tool also exhibited variation of the contact areas as patellar and tibial cartilage mesh material points. In Chapter 6, PLSR knee models were developed, demonstrating the ability to characterise complex morphologic and kinematics interactions and present them in a manner that provides comprehensive clinical insights and applications. The PLSR model closely produced weight-bearing TF and PF alignments that were previously estimated and reported in Chapter 5. This integrated model could be used for rapid examination of the knee joint kinematics in a clinical motion analysis system. Future work will combine the FE knee modelling results with a rigid body model of the musculoskeletal system using open source modelling software (OpenSim, SimTK.org).