Person-specific movement signatures through low profile acquisition and advanced analytics

The field of instrumented gait analysis is changing rapidly with innovations in inexpensive and scalable technologies for remote data capture and advancements in data and computer science. Traditional laboratory-based biomechanics assessments provide high-fidelity data, but require substantive calibration, participant preparation with subjective placement of markers, and therefore have limited accessibility for diverse population study, and limited ability to provide large volumes of data reflective of natural, built, and social environments. This is a major barrier to big data applications, uptake, and innovation. The field has therefore seen significant uptake of more accessible motion capture techniques, including markerless (video-based) motion capture and sensor systems, most notably inertial measurement units (IMUs). The uptake of these technologies has been exponential in recent years across the digital health and behavior change markets, however academic advancements that bridge high-fidelity outcomes with more accessible capture are critically needed. Redundancy in the human musculoskeletal system allows for considerable variability among people, and modeling person-specific biomechanics can provide insight into motor control, dimensionality and plasticity, and person-specific diagnoses or performance indicators. The first arm of this discovery research is focused on leveraging accessible technologies for developing algorithms for person-specific biomechanics signature envelopes from kinematic data. This will build from my previous discovery research including advanced data processing and multivariate modeling, with further advancement with analytics to predict these outcomes based on inertial sensor data. Validity and reliability of IMUs and video-based motion capture for gait outcomes has been limited primarily to kinematic outcomes, however capture of interjoint kinetic outcomes is essential for applications that require an understanding of the underlying neural control and mechanical adaptations. The second arm of this research will aim to develop and utilize algorithms for the valid and reliable representation joint kinetics during gait through wearable sensors and video-based motion capture. Kinetics outcomes will be based on previous algorithm development for capturing amplitude, temporal patterns and frequency of joint forces, with the application of advanced analytics. This discovery research program will build on my decades of previous research into modeling human gait variability in laboratory-based protocols and my transition to community-based motion capture protocols, developing analytical techniques for the identification of deviations in gait kinematics and kinetics related to musculoskeletal pathology and injury. Through this research, I will look forward to continuing to offer an excellent, inclusive environment for training a diverse cohort of highly qualified individuals in this multidisciplinary field.