20 September 2024 in IEEE Transactions on Biomedical Engineering
Objective: Our aim is to determine if data collected with inertial measurement units (IMUs) during steady-state running could be used to estimate ground reaction forces (GRFs) and to derive biomechanical variables (e.g., contact time, impulse, change in velocity) using lightweight machine-learning approaches. In contrast, state-of-the-art estimation using LSTMs suffers from prohibitive inference times on edge devices, requires expensive training and hyperparameter optimization, and results in black box models. Methods: We proposed a novel lightweight solution, SVD Embedding Regression (SER), using linear regression between SVD embeddings of IMU data and GRF data. We also compared lightweight solutions including SER and k-Nearest-Neighbors (KNN) regression with state-of-the-art LSTMs. Results: We performed extensive experiments to evaluate these techniques under multiple scenarios and combinations of IMU signals and quantified estimation errors for predicting GRFs and biomechanical variables. We did this using training data from different athletes, from the same athlete, or both, and we explored the use of acceleration and angular velocity data from sensors at different locations (sacrum and shanks). Conclusion: Our results illustrated that lightweight solutions such as SER and KNN can be similarly accurate or more accurate than LSTMs. The use of personal data reduced estimation errors of all methods, particularly for most biomechanical variables (as compared to GRFs); moreover, this gain was more pronounced in the lightweight methods. Significance: The study of GRFs is used to characterize the mechanical loading experienced by individuals in movements such as running, which is clinically applicable to identify athletes at risk for stress-related injuries.
1 February 2022 in Computers in Biology and Medicine
1 June 2025 in Journal of Applied Biomechanics
The incidence of lower extremity injuries in collegiate distance runners is ∼20%. Identification of a runner sustaining a potential injury remains challenging. This exploratory, cross-institutional study sought to determine whether ground reaction force (GRF) characteristics during steady-state running could identify competitive collegiate distance runners who would later sustain lower extremity injuries. Normative boundaries for 10 GRF variables during braking and propulsion were established for noninjured runners using median ± scaled median absolute deviation. A subanalysis was conducted on runners with and without impact peaks in vertical GRF to mitigate the influence of impact peaks on GRF variables. We hypothesized that prior to injury, runners who later developed an injury would have more GRF variables outside of the normative boundaries than noninjured runners. Using Cliff’s method, a rank-based, nonparametric method for comparing 2 independent groups, we found no statistically significant difference between the number of variables outside the boundaries for injured and noninjured runners overall (P = .17). However, injured runners without impact peaks had more variables outside the normative boundaries than noninjured runners (P < .001). This novel analytical approach demonstrates the potential for preidentifying collegiate distance runners without impact peaks who may be at risk for injury.
1 January 2026 in Journal of Applied Biomechanics
Sprinters need to generate horizontal impulse quickly in the first step out of the blocks to accelerate the body toward the finish line. The purpose of this study was to determine how body configuration at initial foot contact was correlated to ground reaction force measures and multijoint control during the first step out of the blocks by highly trained to world-class sprinters. Measurements of ground reaction forces and segment kinematics during sprint starts performed during a training session revealed that positioning the foot further behind the center of mass at initial contact significantly correlated with shorter contact times, greater average horizontal forces, and increased net joint moment impulse on the knee during the impact phase. Increases in average horizontal forces were significantly correlated with smaller magnitudes of shank angular velocity during the impact phase and shorter times to peak thigh angular velocity during the postimpact phase. We also used kinetic and kinematic data to provide feedback regarding an athlete’s mechanics to coaches for use within a training session on the track.