• Builder
• foot
• gait analysis
• kinematics
• musculoskeletal
• repeatability

This thesis is the result of a five month internship carried out at the Mechanical Engineering Department of The University of Sheffield in the framework of the MD-Paedigree project.
The purpose of the work was to test the repeatability of a previously reported modelling pipeline for the creation of patient-specific musculoskeletal models. It involved the use of motion capture data, MRI images and joint kinematics information. The subject-specific portion of the model was the foot and the ankle, whereas the leg and shank were derived from a generic scaled model. The study included three operators and three patients affected by Juvenile Idiopathic Arthritis (JIA), which is a pediatric pathology of unknown origin. The disease causes ankle joint deformities and pain that cause gait abnormalities. Thus, the need to create patient-specific musculoskeletal models is evident.
Several statistical tests were performed to assess whether operator dependent procedures (i.e. virtual palpation) were influencing the repeatability of the model in terms of kinematic output. A highly repeatable atlas of anatomical markers to be palpated on bone segmentations was identified. However, the virtual palpation of muscular attachments performed according to MRI images was found to be dependent on the operator. It is suggested that skilled experts are desirable at that stage of the pipeline. Kinematics output (i.e. ankle joint angle) was most affected by offset coming from two operator dependent procedures: the registration of the generic scaled model onto the patient-specific model and the creation of joint reference frames. The model allows for inverse dynamics simulation to quantify the influence of operator related procedures on calculated muscles forces and moment arms, being the most sensitive output of musculoskeleta models. It also has the potential to evolve into a three-segment foot to discriminate between loads on different segments of the foot.