The purpose of this thesis was to investigate gait analysis methods used to minimise soft tissue artefact (STA) when measuring the kinematics and kinetics of human gait. STA affects all gait analysis methods and is considered to be one of the major sources of error in clinical motion analysis. A systematic review was initially performed to quantify the magnitude of STA for different areas of the lower limbs during gait analysis. Despite the high quality of the existing literature, the results of the review were inconclusive regarding the exact magnitude of STA during human gait analysis. Previously, there were no methods used consistently throughout the reviewed studies to assess STA. The primary aim of this thesis was to determine the most valid method for minimising STA during human gait analysis. / Whilst the systematic review provided equivocal results on the magnitude of STA, it was able to confirm that STA at the tibia is less than for the femur segment. As a result, the tibial segment was investigated to determine marker locations that are least susceptible to STA. Twenty unimpaired young adults were included in the study and were instrumented with 36 markers, including 10 markers on each shank segment. The markers were well spread across the tibial segment in order to assess as many locations as possible. Four markers located on the tibia were less susceptible to STA. These were the proximal and distal anterior tibial crest markers as well as the medial and lateral malleolar markers. These markers were rigid to one another thus were rigid to the underlying bone. / In order to assess the modelling methods proposed in the literature a gold standard comparison was required. A potential new gold standard method was identified as 3-D freehand ultrasound (3-DUS). This was believed to be a non-invasive and cost effective method for locating internal bony structures. A validation of the new method (3-DUS) against MRI was performed to ensure the new gold standard was a valid methodology. The two methods, 3-DUS and MRI were compared for their accuracy in determining the location of the HJC within the pelvis segment. Twenty unimpaired participants were included in this study. The participants were of variable ages and physical composition. The difference between 3-DUS and MRI determined distance between the left and right HJC was 4.0 ± 2.3mm. It was determined that the results from 3-DUS were clinically not significantly different to MRI. The results of this investigation indicated that 3-DUS could be used as a gold standard measurement for three dimensional gait analysis (3-DGA) research. / The new gold standard method was used to validate existing 3-DGA modelling methods to determine which obtained the most accurate location of the HJC. To date, the greatest clinical application for gait analysis is as a test for people with central nervous system disorders associated with spasticity, especially children with cerebral palsy (CP) (Simon 2004). For this reason, 53 patients with gait abnormalities who had been referred to the Royal Children’s Hospital gait laboratory for a 3-DGA were tested. The participant sample represented patients who were referred to the Hugh Williamson Gait Analysis Laboratory (HWGAL). This was apparent because from 2008-2009 69% of patients at HWGAL had a diagnosis of CP, of the sample included in this study, 67% had a CP diagnosis. / Patients underwent a 3-DGA in addition to a 3-DUS of their left and right femoral heads. Resultant ultrasounds were assessed for the quality of the images and 46 patients were included for data analysis. Seven different methods were investigated for the determination of the HJC and four of these were analysed in two different ways, as such there were 11 models compared to 3-DUS. The Harrington et al method obtained the most accurate and repeatable results where the 3-D location error was 14.3 ± 8.0mm. That method considerably outperformed the functional techniques that had previously been proposed in the literature. This highlighted the importance of testing research techniques in target populations. / To conclude, this thesis has identified locations on the tibia which are most rigid to the underlying bone as well as a new gold standard measurement tool suitable for use in 3-D gait analysis research. The thesis has also demonstrated the validity of using functional methods for determining the HJC in pathological populations. Limitations of previous research were identified, including a lack of translation of research findings into clinical practice. Future work following on from this thesis should aim to address this issue.

Les Artefacts de Tissus Mous (ATM) sont actuellement une des limitations principales pour la mesure du mouvement du membre supérieur avec les techniques actuelles d'analyse du mouvement. L'optimisation multi-segmentaire (OMS) a déjà prouvé son efficacité pour la mesure du mouvement du membre inférieur. Afin d'avoir la meilleure correction possible, il est nécessaire d'utiliser des modèles d'articulation proches de l'anatomie. L'objectif de cette thèse a donc été de développer et de valider un modèle du membre supérieur qui pourrait être utilisé pour la correction des ATM par OMS. De nouveaux modèles en boucle fermée de l'avant-bras et de la ceinture scapulaire ont ainsi été développés accompagnés d'un nouveau modèle de l'articulation scapulo-thoracique imposant à la scapula d'être tangente à un ellipsoïde modélisant le thorax. Ces nouveaux modèles ont été confrontés aux modèles courants de la littérature à travers une étude avec vis intra-corticales sur cadavre et in vivo sur sujets asymptomatiques. Des niveaux d'erreur similaires ont été observés pour tous les modèles quant à leur capacité de corriger les ATM et d'imiter la cinématique osseuse. Les nouveaux modèles semblent cependant beaucoup plus intéressants dans une perspective de développement d'un modèle musculo- squelettique. En effet, le modèle d'avant-bras autorise à la fois d'avoir le mouvement du radius et de l'ulna tandis que le modèle scapulo-thoracique représente mieux la contrainte existant entre le thorax et la scapula. En résumé, cette thèse a permis de développer un modèle complet proche de l'anatomie du membre supérieur permettant de corriger les ATM en utilisant une OMS. Bien que la correction des ATM obtenue n'est pas aussi satisfaisante qu'espérée, l'utilisation de cette approche pour le développement de futurs modèles musculo-squelettique a été validée / Soft Tissue Artefact (STA) is one of the most important limitations when measuring upper limb kinematics through marker-based motion capture techniques, especially for the scapula. Multi Body Optimisation (MBO) has already been proposed to correct STA when measuring lower limb kinematics and can be easily adapted for upper limb. For this purpose, the joint kinematic constraints should be as anatomical as possible. The aim of this thesis was thus to define and validate an anatomical upper limb kinematic model that could be used both to correct STA through the use of MBO and for future musculoskeletal models developments. For this purpose, a model integrating closed loop models of the forearm and of the scapula belt have been developed, including a new anatomical-based model of the scapulothoracic joint. This model constrained the scapula plane to be tangent to an ellipsoid modelling the thorax. All these models were confronted to typical models extracted from the literature through cadaveric and in vivo intracortical pins studies. All models generated similar error when evaluating their ability to mimic the bones kinematics and to correct STA. However, the new forearm and scapulothoracic models were more interesting when considering further musculoskeletal developments: The forearm model allows considering both the ulna and the radius and the scapulothoracic model better represents the constraint existing between the thorax and the scapula. This thesis allowed developing a complete anatomical upper limb kinematic chain. Although the STA correction obtained was not as good as expected, the use of this approach for a future musculoskeletal models has been validated

L'analyse du mouvement humain s'appuie généralement sur des techniques de suivi de marqueurs cutanés pour reconstruire la cinématique articulaire. Cependant, ces techniques d'acquisition présentent d'importantes limites dont les " artefacts de tissus mous " (i.e., le mouvement relatif entre les marqueurs cutanés et le squelette sous-jacent). La méthode d'optimisation multi-segmentaire viseà compenser ces artefacts en imposant aux trajectoires de marqueurs les degrés de liberté d'un modèle cinématique prédéfini. Les liaisons mécaniques modélisant classiquement les articulations empêchent toutefois une estimation satisfaisante de la cinématique articulaire. Cette thèse aborde des perspectives d'amélioration de la méthode d'optimisation multi-segmentaire pour l'estimation de la cinématique articulaire du membre inférieur,à travers différentes approches : (1) la reconstruction de la cinématique par suivi de la vitesse angulaire, de l'accélération et de l'orientation de centrales inertiellesà la place du suivi de marqueurs, (2) l'introduction d'un modèle articulaire élastique basé sur la matrice de raideur du genou, permettant une estimation physiologique de la cinématique articulaire et (3) l'introduction d'un modèle des artefacts de tissus mous " cinématique-dépendant ", visantà évaluer et compenser les artefacts de tissus mous simultanément avec l'estimation la cinématique articulaire. Ce travail a démontré la polyvalence de la méthode d'optimisation multi-segmentaire. Les résultats obtenus laissent espérer une amélioration significative de cette méthode qui devient de plus en plus utilisée en biomécanique, en particulier pour la modélisation musculo-squelettique / Human movement analysis generally relies on skin markers monitoring techniques to reconstruct the joint kinematics. However, these acquisition techniques have important limitations including the "soft tissue artefacts" (i.e., the relative movement between the skin markers and the underlying bones). The multi-body optimization method aims to compensate for these artefacts by imposing the degrees of freedom from a predefined kinematic model to markers trajectories. The mechanical linkages typically used for modeling the joints however prevent a satisfactory estimate of the joint kinematics. This thesis addresses the prospects of improvement of the multi-body optimization method for the estimation of joint kinematics of the lower limb through different approaches: (1) the reconstruction of the kinematics by monitoring the angular velocity, the acceleration and the orientation of magneto-inertial measurement units instead of tracking markers, (2) the introduction of an elastic joint model based on the knee stiffness matrix, enabling a physiological estimation of joint kinematics and (3) the introduction of a "kinematic-dependent" soft tissue artefact model to assess and compensate for soft tissue artefact concurrently with estimating the joint kinematics. This work demonstrated the versatility of the multi-body optimization method. The results give hope for significant improvement in this method which is becoming increasingly used in biomechanics, especially for musculoskeletal modeling

Lors de l'analyse 3D du mouvement humain basée sur des marqueurs cutanés, la position des os ne peut être qu'indirectement estimée. Au cours d'une tâche, les déformations des tissus mous génèrent des déplacements des marqueurs par rapport à l'os : les artefacts de tissus mous (STA), entraînant des effets dévastateurs sur l'estimation de la position. La compensation des STA demeure une question ouverte. L'objectif de cette thèse est de contribuer à la solution de cette question cruciale. La modélisation des STA en utilisant des variables spécifiques mesurables est une condition préalable à son élimination. Un modèle corrigeant les trajectoires individuelles de marqueurs de la cuisse, calibré par des mesures directes des STA, est d'abord présenté. Les STA sont modélisés comme une combinaison linéaire des angles articulaires impliqués. Trois représentations des STA par une série de modes sont proposées : déplacements de marqueurs individuels, transformations géométriques de clusters de marqueurs (MCGT), et variations de forme de l'enveloppe de peau. Le MCGT permet de dissocier les composantes rigides et non rigides. Il a été démontré que seule la composante rigide affecte la cinématique articulaire. Un modèle de cette composante est alors défini pour les clusters cuisse et jambe. Un compromis acceptable entre la correction des STA et le nombre de paramètres a ainsi été obtenu. Les principales applications sont de générer une simulation réaliste des STA ; et surtout, en se concentrant sur la composante rigide, le modèle permet une reconstruction satisfaisante des STA avec moins de paramètres, ce qui facilite son incorporation dans un algorithme d'estimation de la position osseuse / In 3D human movement analysis performed using stereophotogrammetry and skin markers, bone pose can be only indirectly estimated. During a task, soft tissue deformations make the markers move with respect to the underlying bone generating soft tissue artefacts (STA), causing devastating effects on pose estimation and its compensation remains an open issue. The thesis’ aim was to contribute to the solution of this crucial issue. Modelling STA using measurable trial-specific variables is a prerequisite for its removal from marker trajectories. Two STA model architectures are proposed. A thigh marker-level model is first presented. STA was modeled as a linear combination of joint angles involved in the task. The model was calibrated with direct STA measures. The considerable number of model parameters led to defining STA approximations. Three definitions were proposed to represent STA as series of modes : individual marker displacements, marker-cluster geometrical transformations (MCGT), and skin envelope shape variations. Modes were selected using two criteria : modal energy and selecting them a priori. The MCGT allows to select either rigid or non-rigid components. It was also demonstrated that only the rigid component affects joint kinematics. A model of thigh and shank rigid component at cluster-level was then defined. An acceptable trade-off between STA compensation and number of parameters was obtained. These results lead to two main potential applications : generate realistic STAs for simulation