Artificial intelligence (AI) algorithms for gait analysis rely on properly identified foot strikes for step-based feature calculation. Smartphone signals collected during movement assessments, such as the 6-minute walk test (6MWT), have been used to train AI models for foot strike identification and fall risk classification in able-bodied populations. However, there is limited research in populations with more asymmetrical gait. People with lower limb amputation can have high gait variability, adversely affecting automatic step detection algorithms. Hence, fall risk models for lower limb amputees have relied on manual foot strike labelling to calculate step-based features for model training, which is inefficient and impractical for clinical use. In this thesis, decision tree and long-short term memory (LSTM) models were developed, optimized, and their performance compared for automated foot strike identification in an amputee population. Eighty people with lower limb amputations (27 fallers, 53 non-fallers) completed a 6MWT with a smartphone at the posterior pelvis. Automated and manually labelled foot strikes from the full 6MWT and from the first two minutes of data were used to calculate step-based features. A random forest model was used to classify fall risk. The best foot strike identification model was an LSTM with 100 hidden nodes in the LSTM layer, 50 hidden nodes in the dense layer, and batch size of 64 (99.0% accuracy, 86.4% sensitivity, 99.4% specificity, 82.7% precision). Automated foot strikes from the full 6MWT data correctly classified more fallers (55.6% versus 48.1%), whereas automated foot strikes from 2-minute data classified more non-fallers (90.6% versus 81.1%). Feature calculation using manually labelled foot strikes resulted in the best overall performance (80.0% accuracy, 55.6% sensitivity, 92.5% specificity). This research created a novel method for automated foot strike identification in lower limb amputees that is equivalent to manual labelling and demonstrated that automated foot strikes can be used to calculate step-based features for fall risk classification. Integration of the foot strike identification model into a smartphone application could allow for immediate stride analysis after completing a 6MWT; however, fall risk classification model improvement is recommended to enhance clinical viability.

No / The purpose of this article is determining if and how the kinematic relationship between adjacent body segments changes when an individual’s gait pattern is experimentally manipulated can yield insight into the robustness of the kinematic coupling across the associated joint(s). The aim of this study was to assess the effects on the kinematic coupling between the forefoot, rearfoot and shank during ground contact of running with alteration in foot strike pattern. Twelve subjects ran over-ground using three different foot strike patterns (heel strike, forefoot strike, toe running). Kinematic data were collected of the forefoot, rearfoot and shank, which were modelled as rigid segments. Coupling at the ankle-complex and midfoot joints was assessed using cross-correlation and vector coding techniques. In general good coupling was found between rearfoot frontal plane motion and transverse plane shank rotation regardless of foot strike pattern. Forefoot motion was also strongly coupled with rearfoot frontal plane motion. Subtle differences were noted in the amount of rearfoot eversion transferred into shank internal rotation in the first 10–15% of stance during heel strike running compared to forefoot and toe running, and this was accompanied by small alterations in forefoot kinematics. These findings indicate that during ground contact in running there is strong coupling between the rearfoot and shank via the action of the joints in the ankle-complex. In addition, there was good coupling of both sagittal and transverse plane forefoot with rearfoot frontal plane motion via the action of the midfoot joints.

One major way that running style varies between individuals is how their foot strikes the ground (forefoot strike or rearfoot strike). Running generates a torque about the ankle that depends on the individual’s foot strike pattern, length of their plantar flexor moment arm, and force generated from the plantar flexor muscles. The foot strike pattern during running, gastrocnemii muscle thickness, and heel length (used as an approximation for plantar flexor moment arm) were determined in 41 runners. Forefoot and rearfoot strike runners had the same thickness of the gastrocnemii muscles. However, in comparison with sedentary walkers, the runners had thicker calf muscles. These results imply a “peak” muscle thickness seems to be attained by running. Runners with longer heels were more likely to use a forefoot strike running style, possibly due to a mechanical advantage in the generation of torque.

Mathematical modeling and realistic human simulation of human jogging motion is a very challenging problem. Majority of the current literature is focused on studying walking or running. This work is aimed at bridging the gap in literature due to the lack of research work in three main areas: (1) simulations and experiments on running at speeds lower than 3 m/s, (2) Kinetics of fore-foot strike pattern in jogging and running and (3) the existence of a double support phase in running at slower speeds and its effects. Formulations to simulate natural human jogging are studied and developed. The digital human model used for this work includes 55 degrees of freedom, 6 for global translation and rotation and 49 for the revolute joints to represent the kinematics of the body. Predictive Dynamics methodology is used for dynamic analysis where the problem is formulated as a nonlinear optimization problem. Both, displacement and forces are considered as unknowns and identified by solving the optimization problem. The equations of motion are satisfied by applying them as equality constraints in the formulation. Kinematics analysis of the mechanical system is performed using the Denavit-Haretneberg (DH) method. The zero moment point (ZMP) condition is satisfied during the ground contact phase to achieve dynamic stability. The joint angle profiles are discretized using B-spline interpolation method. The joint torque squared, also termed dynamic effort, and the difference between predicted motion and motion capture data are used as performance measures and minimized in the optimization formulation. The formulation also includes a set of constraints to simulate natural jogging motion. Two formulations are discussed for jogging on a straight path: (1) one-step jogging formulation and (2) one-stride jogging formulation. The one-stride formulation is discussed for clock-wise and counter clock-wise jogging along a curved path. Cause and effect is shown by obtaining simulation results for different loading conditions. The proposed formulation provides realistic human jogging motion and is very robust.

The main goal of the project was to compare how the Achilles tendon is affected while running with traditional shoes, minimalistic shoes and barefoot. Displacement and strain were calculated both for different shoes and for different foot strike patterns. The calculations were done with a speckle tracking algorithm and displacement was calculated for three different depths in the tendon: deep layer, mid layer and superficial layer. The goal was also to conduct this analysis after a review of the algorithm used. The review of the algorithm focused on the size of the region of interest, kernel size and frequency. Literature study showed that it is more common to use a smaller kernel size, but the same shape. The region of interest was chosen depending on the size of the tendon. Displacement and strain in the Achilles tendon was calculated for seven subjects and the result did not show any difference in amount of mean deformation due to different shoe types or foot strike patterns. It was a small sample group but the result indicated a difference in peak displacement between deep and superficial layer depending on different shoe types and foot strike patterns. The difference in peak displacement between deep and superficial layer was lowest when running barefoot, larger when running with minimalistic shoes and greatest when running with traditional shoes. This result was only achieved when running with rear foot strike pattern. When running with fore foot strike pattern the difference in peak displacement between layers did not change with different conditions. In all conditions the difference in peak displacement between the layers was greater when running with rear foot strike pattern than when running with front foot strike pattern. The deep layer displaced more than the superficial layer (p<0.01) for all conditions and foot strike patterns.

Accelerometers and gyroscopes are used to detect foot strike (FS), i.e., the moment when the foot first touches the ground. However, it is unclear whether different conditions (footwear hardness or foot strike pattern) influence the accuracy and precision of different FS detection methods when using such micro-electromechanical sensors (MEMS). This study compared the accuracy of four published MEMS-based FS detection methods with each other and the gold standard (force plate) to establish the most accurate method with regard to different foot strike patterns and footwear conditions. Twenty-three recreational runners (12 rearfoot and 11 forefoot strikers) ran on a 15-m indoor track at their individual running speed in three footwear conditions (low to high hardness). MEMS and a force plate were sampled at a rate of 3750 Hz. Individual accuracy and precision of FS detection methods were found which were dependent on running styles and footwear conditions. Most of the methods were characterized by a delay which generally increased from rearfoot to forefoot strike pattern and from high to low midsole hardness. It can be concluded that only one of the four methods can accurately determine FS in a variety of conditions.

Bien que constitué anatomiquement et physiologiquement pour la course d’endurance, l’Homme est considérablement exposés à diverses blessures musculo-squelettiques liées à la répétition de contraintes mécaniques. Le coureur de trail running par exemple est soumis à de nombreux impacts ainsi qu’à une fatigue et des dommages musculaires sévères. Ces chocs répétitifs et dommages musculaires réduiraient la tolérance du coureur face aux contraintes mécaniques le poussant ainsi à altérer sa cinématique de course. Par conséquent, minimiser les dommages musculo-squelettiques serait déterminant pour la performance en trail running. Des évidences montrent que la pose de pied altère la localisation et l’intensité des contraintes appliquées au système musculo-squelettique. L’objectif de ce travail de thèse a été d’étudier l’influence du pattern de pose de pied sur l’impact et la fatigue neuromusculaire en trail running. Les phases de descente ont été tout particulièrement étudiées du fait qu’elles sont les plus traumatisantes. En effet, ce travail de thèse a mis en évidence qu’en situation de trail running, l’intensité de l’impact augmente lorsque la pente diminue, et que la fatigue neuromusculaire périphérique est aussi sévère à la suite d’une descente isolée qu’après un ultra-trail de plusieurs heures. En étudiant l’influence de la pose de pied adoptée au cours d’une descente en situation de trail running, il a été observé qu’attaquer le sol par l’avant du pied augmentait la fatigue neuromusculaire aux extenseurs du genou. Cependant, une importante variabilité dans les patterns de pose de pied adoptés au cours de la descente a été associée à une baisse de la fatigue neuromusculaire aux extenseurs du genou et fléchisseurs plantaires. L’influence de la pose de pied sur l’intensité du choc et le contenu vibratoire le long des axes axial et transversal a également été démontrée : adopter une attaque talon diminue la sévérité du choc axial mais réduit l’intensité du choc transversal. La principale conclusion est qu’aucun pattern de course ne saurait être universellement recommandé du fait que « changer de pose de pied » est synonyme de « changer la localisation et la magnitude des contraintes appliquées au système musculo-squelettique ». En ce sens, alterner entre différents patterns de course serait une stratégie efficiente en trail running / Although Humans are “born” anatomically and physiologically adapted to long distances run, they are substantially exposed to various musculoskeletal overuse injuries. Trail runners sustain a high number of foot-to-ground contacts and develop severe muscle fatigue and damages. Repetitive shocks and muscle damages would reduce the runners’ tolerance to mechanical strains leading to changes in running kinematics. Minimizing musculoskeletal damages is therefore considered paramount for performance in trail running. Numerous studies highlighted that the foot strike pattern alters the localization and magnitude of the mechanical strains applied on the musculoskeletal system. The main purpose of this thesis was to study the influence of the foot strike pattern on impact and neuromuscular fatigue in trail running. Downhill sections were mainly investigated since they are the most mechanically stressful. Indeed, it was observed from this thesis’ work that, in real trail running practice, the impact intensity increases as the slope decreases, and that the neuromuscular fatigue induced by a single downhill run is as severe as the one induced by an ultratrail race that lasts several hours. Investigating the effect of the foot strike pattern adopted during a downhill trail run on fatigue, it was observed that forefoot striking increases the neuromuscular fatigue at knee extensors. However, a high variability in foot strike patterns adopted was associated to a lower neuromuscular fatigue at both knee extensors and plantar flexors. The effect of the foot strike pattern on axial and transversal shock and vibration content was also demonstrated: heel striking was correlated to a lower impact severity along the axial axis of the skeleton but a greater one along its transversal axis. The main conclusion of this thesis is that no single foot strike pattern should be universally advised due to “changing of foot strike” means “changing the localization and magnitude of the mechanical stress applied on the musculoskeletal system”. Switching between different running patterns might be an efficient strategy in trail running

Accelerometers and gyroscopes are used to detect foot strike (FS), i.e., the moment when the foot first touches the ground. However, it is unclear whether different conditions (footwear hardness or foot strike pattern) influence the accuracy and precision of different FS detection methods when using such micro-electromechanical sensors (MEMS). This study compared the accuracy of four published MEMS-based FS detection methods with each other and the gold standard (force plate) to establish the most accurate method with regard to different foot strike patterns and footwear conditions. Twenty-three recreational runners (12 rearfoot and 11 forefoot strikers) ran on a 15-m indoor track at their individual running speed in three footwear conditions (low to high hardness). MEMS and a force plate were sampled at a rate of 3750 Hz. Individual accuracy and precision of FS detection methods were found which were dependent on running styles and footwear conditions. Most of the methods were characterized by a delay which generally increased from rearfoot to forefoot strike pattern and from high to low midsole hardness. It can be concluded that only one of the four methods can accurately determine FS in a variety of conditions.