Personalization of generic musculoskeletal model in OpenSim via MRI imaging for individualized gait analysis

The present thesis is the product of the work done by the candidate with the support offered by the supervisor and the co-supervisors during a six months period thanks to the collaboration between the Human Movement Bioengineering Laboratory Research group of the University of Padova (Italy) and the Human Movement Biomechanics Research group at KU Leuven (Belgium). Diabetes and diabetes related pathologies affect every year millions of people around the globe. One of the worst consequences related to diabetes which may develop when the problem is left untreated is diabetic neuropathy: this pathology affects the lower limb nerves and vascular system. The most common consequences are: alteration of the peripheral sensory system, with consequent lack of protective sensation on the sole of the foot, muscle weakening and thickening of the tendons and ligaments of the foot, cartilage tissue stiffening with consequent reduction in the ankle and foot joints range of motion. Furthermore foot deformities may develop, generating non-physiological pressure levels on some area of the plantar aspect of the foot: this can lead to the formation of calluses and of ulcers on the foot sole. In case the ulcers are left untreated, this may lead to amputation of the foot. In this context the study of foot biomechanics has shown to play a substantial role in preventing foot ulcers on the feet of diabetic subjects. In the literature there is a wide variety of foot musculoskeletal models [63,80], in particular some are specifically devised for allowing the comparison between healthy subjects and subjects affected by the diabetic foot pathology. The aim of the present project is to develop a subject-specific musculoskeletal model able to simulate the mechanics of the foot during the gait cycle movements, based on MRI images combined with gait analysis data, to be applied in the context of diabetic foot prevention. Subject-specific modeling is essential to determine variables that cannot be directly measured, such as muscle forces, ligament elongations and so on. In the context of foot modeling, the simulation of gait can provide insights about the ongoing process occurring at the foot bones and on the soft tissues, like tendons and ligaments. Foot musculoskeletal modeling techniques can be very useful in understanding why and where foot ulcers develop [67] and in the assessment of surgical intervention outcomes or in prescribing plantar orthosis for ulcer prevention [79]. The current project was organized as follows: first a procedure was developed in order to personalize through subject specific MRI scans a generic state of the art foot model developed by Malaquias et al [63], the so called extended foot model in OpenSim; second the generic model local embedded coordinate systems was updated by implementing Padova Foot Model by Sawacha et al [78] and applying these anatomical embedded system to the extended foot model [63]; hence the orientation of the body parts of the personalized foot model within the extended foot model [63] in OpenSim was defined according to this new reference system [78]; third the results of the gait simulations obtained through the personalized model and the generic one were compared in terms of: • Inverse Kinematics (IK) • Inverse Dynamics (ID) • Static Optimization (SO) Finally a further comparison was carried on between the results obtained from the IK tool (using the MRI specific model where the foot local body frames have been defined following the Padova Foot Model) and the direct kinematics results obtained applying the Padova Foot Model. The thesis is articulated in 5 chapters: the first one introduces the diabetes pathology. The second chapter describes movement analysis.The third chapter describes the musculoskeletal modeling characteristics. The 4th describes the materials and methods followed, the 5th is the results' chapter.