This work demonstrates an application of Vibration-Based Occupant Inference (VBOI) in frailty analysis. The rise of both Internet-of-Things (IoT) and VBOI provide new techniques to perform gait analysis via footstep-induced vibration which can be analyzed for early detection of human frailty. Thus, this work provides an application of VBOI to passively track gait parameters (e.g., gait speed) using floor-mounted accelerometers as opposed to using a manual chronometer as it is commonly performed by healthcare professionals. The first part of this thesis describes the techniques used for footstep detection by measuring the power of the footstep-generated vibration waves. The extraction of temporal gait parameters from consecutive footsteps can then be used to estimate temporal features such as cadence and stride time variation. VBOI provides many algorithms to accurately detect when a human-induced vibration event happened, however, spatial information is also needed for many gait parameters used in frailty diagnosis. Detecting where an event happened is a complicated problem because footsteps waves travel and decay in different ways according to the medium (floor system), the number of people walking, and even the walking speed. Therefore, the second part of this work will utilize an energy-based approach of footstep localization in which it is assumed that footstep waves decay exponentially as they travel across the medium. The results from this approach are then used to calculate spatial and tempo-spatial parameters. The main goal of this study is to understand the applicability of VBOI algorithms in gait analysis for frailty detection in a healthcare setting. / Master of Science / Human frailty is responsible for one of the highest healthcare costs and the death of many people every year. Although anyone suffering from frailty has a higher chance of death, it is particularly dangerous for the elderly population and for those suffering from other comorbidities. Diagnosing frailty is hard because it usually happens slowly over time. However, it has been shown that changes in some walking parameters (such as gait speed) can be an early indication of frailty. Many technologies have been created in order to track gait parameters, many of which either require expensive equipment (e.g., force plates) or the use of wearable devices, which can introduce privacy concerns. It has been proposed in the literature that Vibration-Based Occupant Inference (VBOI) techniques could be used in healthcare applications. Such algorithms measure footstep-induced vibration waves in order to detect and track footsteps. This system can provide several advantages in frailty analysis because of its affordability, ease of use, and little impact on patients' privacy. Therefore, the aim of this study is to understand the applicability of VBOI algorithms in gait analysis for frailty detection to be used in a healthcare setting. This thesis will proceed as follows: 1- The demonstration of an energy-based footstep detection and localization algorithm in VBOI. 2 - The application of such algorithms for gait parameters extraction with simulated frail walkers. 3 - Finally, an analysis of the proposed VBOI techniques for deployment in a real hospital setting.

Residual gait abnormalities are common following lower limb injury and/or stroke and can have several negative impacts on an individual’s life. Without continuous treatment and follow up, individuals can be prone to chronic pain as abnormal gait may lead to non-physiological loading of the musculoskeletal system. The current industry gold standard for diagnosing abnormal gait requires specialty equipment that is generally only available at designated gait facilities. Due to the inaccessibility and high cost associated with these facilities, a wearable SmartSole device to continuously detect abnormal gait was proposed. A previous iteration of the SmartSole was unable to properly detect abnormal gait and also experienced fracturing throughout the 3D printed body. In this present study, sensor placement and material selection were reconsidered to address these limitations. The objective of this study was to determine if a redesigned SmartSole could identify events of abnormal gait through validation and verification testing against the industry standard force plates. In total, 14 participants were selected for gait studies, 7 with pronounced gait abnormalities (e.g. limps), and 7 with physiological gait. Parameters of interest included stance time, gait cycle time, and the ratio of the force magnitudes recorded during heel strike and toe off. Results indicated that the SmartSole was effective at determining overall event timings within the gait cycle, as both stance and cycle time had strong, positive correlations (left stance: r = 0.761, right stance: r = 0.560, left cycle: r = 0.688) with the force plates, with the exception of right foot cycle time. The sole was not effective at measuring actual values of events during gait as there were weak correlations with the force plates. Furthermore, when comparing parameters of interest between the injured and non-injured sides for test participants with gait abnormalities, neither the SmartSole nor the force plates were able to detect significant differences. The inability of the sole to accurately collect force magnitudes or to detect abnormal gait leads to the conclusion that additional sensors may need to be implemented. Future iterations may consider placement of additional sensors to allow for a “fuller picture” and the inclusion of other types of sensors for improved, continuous tracking of gait abnormalities.

Since the 1960s, research into the medical applications of wearable robots has been fueled by a growing need for assistive technologies that can help individuals impacted by musculoskeletal disorders such as sarcopenia independently manage common activities of daily living while maintaining their natural physical capacities. While contemporary research has demonstrated promising developments, the usefulness of exoskeletons in everyday settings remains limited due to design factors that include the limited practicality of existing battery technologies, the need for actuators exhibiting a high output torque-to-weight ratio, a need for modular designs that are minimally disruptive to wearers, and the need for control systems that can actively work in sync with a user. To explore potential solutions to some of these limiting factors, a novel ankle exoskeleton prototype supporting ankle plantarflexion during gait was developed under a design approach that seeks to optimize actuator performance. The actuation system featured in this prototype consists of a custom dual-action linear actuator that can provide mechanical assistance to both ankles via a single BLDC motor and an underlying Bowden cable system. The metric ball screw and BLDC motor implemented in the linear actuator were selectively chosen to minimize the motor torque and current required to assist wearers impacted by a degree of muscle weakness under an assistance-as-needed design paradigm. The prototype additionally features an array of force sensing resistors for tracking gait progression and exploring potential user-based control strategies for synchronizing the exoskeleton actuator with a wearer's gait. Performance analysis for this prototype was conducted with the goal of quantifying the exoskeleton's force output, actuator settling time, and the control system's ability to track gait and identify key events in the gait cycle. The preliminary findings of this experimental analysis support the viability of the actuator's dual-action concept and gait progression tracking system as a starting ground for future developments that build on a similar design optimization approach. / Master of Science / Healthy aging and good physical health are characterized in part by one's ability to self-manage a core set of daily living tasks, one of the most prominent of which is gait. Relative to existing assistive technologies such as wheelchairs, exoskeletons provide the unique benefit of providing active mechanical support while encouraging users to rely on their natural physical capabilities. While recent technological developments in the field of wearable robots show promise, the viability of exoskeletons in an everyday setting remains constrained in part by three underlying design factors: the limited practicality of existing battery technologies, a need for actuators that can satisfactorily balance a high force output with weight, and a need for control strategies that can properly synchronize wearable robots with users. The ankle exoskeleton prototype introduced in this thesis is a portable, energetically autonomous wearable device that supports ankle plantarflexion during the push-off stages of the gait cycle. The design for this prototype seeks to optimize actuator performance and features a novel dual-action linear actuator that provides walking support to both ankles using a single DC motor. The exoskeleton additionally features an array of contact sensors that track the user's progression throughout the gait cycle and allow for the examination of potential control strategies for synchronizing the actuator with the wearer's gait. Performance analysis conducted for this prototype quantifies the exoskeleton's force output, approximates the actuator's settling time between steps, and assesses the control system's ability to track gait and synchronize with a wearer. The findings from these performance evaluation experiments support the viability of the actuator's dual-action concept and gait progression tracker as a foundation for future developments that build on a similar design optimization approach.

This thesis presents the development and implementation of a machine learning prediction model for concurrently aggregating interval linear step distance predictions before future foot placement. Specifically, on-body inertial measurement units consisting of accelerometers, gyroscopes, and magnetometers, through integrated development by Xsens, are used for measuring human walking behavior in real-time. The data collection process involves measuring activity from two subject participants who travel an intended course consisting of flat, stair, and sloped walking elements. This work discusses the formulation of the ensemble machine learning prediction algorithm, real-time application design considerations, feature extraction and selection, and experimental testing under which this system performed several different test case conditions. It was found that the system was able to predict the linear step distances for 47.2% of 1060 steps within 7.6cm accuracy, 67.5% of 1060 steps within 15.2cm accuracy, and 75.8% of 1060 steps within 23cm. For separated flat walking, it was found that 93% of the 1060 steps have less than 25% error, and 75% of the 1060 steps have less than 10% error which is an improvement over the commingled data set. Future applications and work to expand upon from this system are discussed for improving the results discovered from this work. / Master of Science / This thesis presents the development and implementation of a machine learning prediction model for determining the stepping distance of future steps in real-time walking before their placement occurs. Specialized sensor units for measuring human motion activity are worn on the body for collecting and characterizing human walking behavior in real-time. Two subject participants are asked to walk a planned course consisting of flat, stair, and sloped walking elements. This work discusses the prediction algorithm voting scheme, real-time application design considerations, descriptive data elements for the algorithm, and experimental testing under which this system performed several different test case conditions. Detailed experimental tests are concluded in order to fully understand the extent of the system’s performance and the behaviors it exhibits throughout. The approach explored in this work enables researchers and roboticists to develop improvements and construct variations which may become superior to this method.

Motion capture (mocap) systems integrated with force plates and electromyography (EMG) collect detailed kinematic and kinetic data on subjects, including stride length, width, cadence, speed, and other spatiotemporal parameters. These systems allow clinicians and researchers to analyze movements, both cyclic (e.g., walking, running) and non-cyclic (e.g., jumping, falling), which is crucial for understanding movement patterns and identifying abnormalities. Clinical gait analysis, a key application, focuses on detecting musculoskeletal issues and walking impairments. While essential for diagnosing gait disorders and planning interventions, clinical gait analysis faces challenges such as noise, outliers, and marker occlusion in optical motion tracking data, requiring complex post-processing. Additionally, the measurement of ground reaction forces (GRFs) and moments (GRMs) is limited due to the restricted number of force plates. There are also challenges in EMG data collection, such as finding optimal MVC positions and developing nonlinear normalization techniques to replace traditional methods.To address these challenges, this research aims to develop an AI-driven gait analysis system that is cost-effective, user-independent, and relies solely on kinematic and EMG data for real-time analysis. The system is specifically designed to assess musculoskeletal characteristics in individuals with special needs, walking disabilities, or injuries, where measuring MVC levels is impractical or unsafe. The research has four main objectives: (1) standardize MVC positions for four lower limb muscles, (2) develop an alternative EMG normalization technique using nonlinear data analysis, (3) create an unsupervised framework using transformers for missing marker recovery without perfect ground-truth data, and (4) generate GRFs, GRMs, and JMs from lower limb kinematics using a 1D-CNN, improving accuracy and efficiency with transfer learning, without requiring force plates. While addressing these challenges, the proposed system aims to minimize user interaction, reduce pre- and post-processing, and lower costs for researchers and clinicians. The designed tool will integrate with existing optical marker-based mocap systems, providing greater flexibility and usability. In educational settings, it will offer students hands-on experience in advanced gait analysis techniques. Economically, widespread adoption of the tool in research and clinical settings will reduce data collection and analysis costs, making advanced gait analysis more accessible. Additionally, this tool can be applied to other fields, such as precision manufacturing, security, and predictive maintenance, where analyzing data can predict failures. Consequently, this research will significantly advance the field of human movement, increasing the volume and quality of research using optical marker-based mocap systems.

KISS (Kinematic Support System) is a locally developed gait analysis system at Middle East Technical University (METU), and the performance of the system was evaluated before as a whole. However, such evaluations do not differentiate between the efficacy of the data acquisition system and the model-based gait analysis methodology. In this thesis, kinematic results of the KISS system will be compared with those of the Ankara University based commercial VICON (Oxford Metrics Ltd., Oxford, UK) system, in view of evaluating the performance of data acquisition system and the gait analysis methodology separately. This study is expected to provide guidelines for future developments on the KISS system.

KISS (Kinematic Support System) is a locally developed gait analysis system at Middle East Technical University (METU), and the performance of the system was evaluated before as a whole. However, such evaluations do not differentiate between the efficacy of the data acquisition system and the model-based gait analysis methodology. In this thesis, kinematic results of the KISS system will be compared with those of the Ankara University based commercial VICON (Oxford Metrics Ltd., Oxford, UK) system, in view of evaluating the performance of data acquisition system and the gait analysis methodology separately. This study is expected to provide guidelines for future developments on the KISS system.

This study concerned the development and validation of a hardware and software biomeasurement system, which was designed to be used by physiotherapists, general practitioners and other healthcare professionals. The purpose of the system is to detect and assess gait deviation in the form of reduced post-operative range of movement (ROM) of the replacement hip joint in total hip arthroplasty (THA) patients. In so doing, the following original work is presented: Production of a wearable, microcontroller-equipped system which was able to wirelessly relay accelerometer sensor data of the subject’s key hip-position parameters to a host computer, which logs the data for later analysis. Development of an artificial neural network is also reported, which was produced to process the sensor data and output assessment of the subject’s hip ROM in the flexion/extension and abduction/adduction rotations (forward and backward swing and outward and inward movement of the hip respectively). The review of literature in the area of biomeasurement devices is also presented. A major data collection was carried out using twenty-one THA patients, where the device output was compared to the output of a Vicon motion analysis system which is considered the ‘gold standard’ in clinical gait analysis. The Vicon system was used to show that the device developed did not itself affect the patient’s hip, knee or ankle gait cycle parameters when in use, and produced measurement of hip flexion/extension and abduction/adduction closely approximating those of the Vicon system. In patients who had gait deviations manifesting in reduced ROM of these hip parameters, it was demonstrated that the device was able to detect and assess the severity of these excursions accurately. The results of the study substantiate that the system developed could be used as an aid for healthcare professionals in the following ways: · To objectively assess gait deviation in the form of reduced flexion/extension and abduction/adduction in the human hip, after replacement, · Monitoring of patient hip ROM post-operatively · Assist in the planning of gait rehabilitation strategies related to these hip parameters.

Due to their prevalence and associated high rehabilitation costs, this thesis aimed to better understand factors influencing the risk of tibial (TSF) and third metatarsal (MT3SF) stress fractures in Royal Marine recruit training. In Study 1, the standard issue combat assault boot and neutral trainer were assessed during running. Running in the boot caused restricted ankle motion, greater forefoot loading, greater ankle stiffness and a more laterally applied horizontal force vector at the instant of peak braking, suggesting that the risk of incurring MT3SF was greater in this condition. In Study 2, bending stresses were modelled along the length of the third metatarsal of five participants, using individual bone geometry and dynamic gait data. Stresses were modelled for running when barefoot, and when shod in the standard issue footwear. Estimated peak bending stresses were significantly greater in the combat assault boot than the gym trainer, predominantly due to increased plantar loading. Individual bone geometry was however dominant in determining peak bending stresses. In Study 3, a large (n=1065) prospective study was conducted to identify differences in baseline characteristics between recruits sustaining a TSF or MT3SF and those who complete training uninjured. Ten TSF and 14 MT3SF cases were compared to 120 uninjured legs. Results suggest that risk of TSF is greater in those recruits with reduced ability to resist loading and attenuate impact during gait. Results for MT3SF suggest that ankle and foot position at touchdown, and the timing and magnitude of forefoot loading, are important factors influencing risk of this injury. The observation of lower age and BMI in both stress fracture groups was linked to lower bone strength and earlier fatigue mechanisms. This thesis has increased the understanding of MT3SF in particular, and provides information on specific factors which may be associated with MT3SF and TSF in RM recruits during basic training.

Patellofemoral pain syndrome (PFPS) is the most common injury in runners. This thesis investigates the theory that prolonged eversion at the rear-foot causes prolonged tibial internal rotation and excessive femoral internal rotation, and predisposes female, distance runners to PFPS. Since eversion occurs at the subtalar joint, the morphology of the talus and calcaneus were also assessed. The study was a case-control investigation between female runners with a history of PFPS (n = 9) and normal controls (n = 10). Gait analysis was used to measure lower-limb joint angles during barefoot, treadmill running. It was hypothesised that runners with PFPS would demonstrate prolonged rearfoot eversion and tibial internal rotation, and increased hip internal rotation compared to normal controls. Computed tomography was used to image the foot and ankle in simulated weight-bearing using a custom-built loading rig. Three-dimensional models of the talus and calcaneus were generated and their shapes were quantified using principal axis lengths and orientations. The results did not support the theory that prolonged eversion and rear-foot structure predispose to PFPS during running, and attributing PFPS to these factors should be done with discretion. However, runners with a history of PFPS exhibited increased rear-foot eversion, reduced rear-foot dorsiflexion and increased knee internal rotation compared to normal controls during running, walking and squatting. Subjects with PFPS also demonstrated increased dorsiflexion at the mid-foot. It was proposed that increased eversion was secondary to reduced rear-foot dorsiflexion as this enabled compensatory dorsiflexion at the mid-tarsal complex. Due to the tight articulation of the ankle mortise, increased knee internal rotation corresponds well with excessive rear-foot eversion. A prospective study is required to establish whether these kinematic alterations are a cause or an effect of PFPS.