Southwest Research Institute (SwRI) and St. Mary’s University have embarked on a pioneering collaboration to revolutionize how metabolic cost predictions are made in biomechanical evaluations. Utilizing SwRI’s advanced markerless motion capture technology, known as ENABLE™, combined with sophisticated musculoskeletal modeling and cutting-edge machine learning techniques, this partnership aims to elevate the accuracy and utility of metabolic energy expenditure estimations across clinical and performance disciplines. Funded by a grant from the St. Mary’s-SwRI Technology & Applied Research (S²TAR) program, this initiative reflects a significant stride toward non-invasive, real-time analysis of human movement inefficiency and rehabilitation outcomes.
Metabolic cost — the quantifiable amount of energy the human body consumes during physical activities such as walking, running, or performing everyday tasks — serves as a critical metric in understanding movement efficiency and muscular demands. Dr. Nicholas Vandenberg, a research engineer at SwRI and co-principal investigator, emphasizes the immense value of reliably estimating this parameter: it offers rehabilitative specialists an objective means to tailor therapies aimed at optimizing patients’ energy expenditures, ultimately facilitating improved mobility and reducing fatigue. By advancing predictive capabilities beyond traditional, marker-based systems, the ENABLE platform promises enhanced precision without the encumbrance and complexity of physical markers.
Central to this project is the innovative ENABLE system, which leverages state-of-the-art computer vision and deep learning algorithms to capture three-dimensional kinematics without intrusive markers. This technology transcends conventional biomechanics tools by enabling seamless, markerless motion capture that integrates biomechanical modeling expertise to generate robust datasets ideal for clinical and sports science applications. The deployment of ENABLE particularly focuses on diverse subject groups, including individuals with below-the-knee amputations, reflecting the commitment to addressing mobility challenges through sophisticated engineering solutions.
St. Mary’s University’s expertise in machine learning, led by Dr. Ricardo Ramirez, complements ENABLE’s technological foundation by developing algorithms that interpret 2D video footage to predict metabolic costs. This project significantly expands upon earlier work by incorporating three-dimensional video analyses, thus enriching data granularity and improving prediction fidelity. The synergistic integration of computer vision data and musculoskeletal simulations permits an unprecedented understanding of muscle-specific energy consumption, facilitating the refinement of both biomechanical models and artificial intelligence approaches.
To ensure scientific rigor, the collaboration employs metabolic carts to obtain direct, real-time measurements of energy expenditure from study participants. Such empirical data serve as the ground truth against which the validity of machine learning predictions is tested. By juxtaposing model-generated estimates with observed metabolic rates, the team aims to iteratively enhance algorithmic accuracy, enabling predictive models to capture nuances in human movement efficiency that have traditionally been elusive.
One of the groundbreaking aspects of this research lies in the use of OpenSim models that incorporate individual muscle fibers and dynamics. Through these sophisticated simulations, researchers can dissect the metabolic cost contributions across distinct muscle groups, offering a detailed biomechanical perspective that informs both clinical and athletic interventions. This granular insight is paramount for devising rehabilitation strategies tailored not only to overall movement patterns but also to specific muscular demands that influence fatigue and injury risk.
The implications of this technology extend well beyond clinical rehabilitation. As ENABLE refines its capabilities, its potential application in sports science could redefine athletic training by enabling coaches and therapists to quantify metabolic loads with heightened accuracy. Through personalized data on muscle efficiency and energy expenditure, performance optimization may be approached with precision previously unattainable, fostering advancements in injury prevention and recovery protocols.
Moreover, the practical advantages of using a markerless system facilitate broader accessibility and scalability in real-world settings. The reduction in setup time, participant discomfort, and equipment expenses opens avenues for wider adoption in outpatient clinics, athletic training facilities, and research laboratories. This democratization of biomechanical assessment tools aligns with broader trends toward wearable and non-invasive health monitoring technologies, signaling a paradigm shift in human performance analytics.
The focus on individuals reliant on prosthetic devices underscores the project’s emphasis on addressing critical gaps in mobility research. By illuminating subtle gait inefficiencies and energy expenditure patterns among prosthesis users, the research offers opportunities to tailor prosthetic design and fitting with unprecedented precision. Such customized approaches have the potential to lessen fatigue, enhance comfort, and improve the overall quality of life for millions of individuals experiencing limb loss.
In addition to clinical populations, the robust data generated through this project could catalyze research into neuromuscular diseases, age-related mobility decline, and occupational biomechanics, where metabolic cost assessment is vital. The integration of ENABLE’s motion capture with computational models and machine learning tools exemplifies how interdisciplinary approaches can unravel complex biological processes and translate them into tangible health solutions.
As the research team continues to iterate on the metabolic cost prediction algorithms, their vision encompasses a comprehensive system capable of evaluating a full spectrum of activities, from simple ambulation to high-intensity exercise. By distributing metabolic cost predictions across muscle groups and functional tasks, the technology aspires to offer insights that inform everything from prosthetic user rehabilitation to elite athlete conditioning.
This venture, supported by an investment of $127,750 from the S²TAR program, represents a forward-looking commitment to bridging engineering ingenuity and biomedical science. ENABLE embodies the next frontier in biomechanical evaluation, merging artificial intelligence and biomechanics for a future where movement efficiency metrics are accessible, reliable, and deeply informative. The outcomes of this partnership hold promise for transforming rehabilitative care, enhancing prosthetic technologies, and optimizing human performance across diverse populations.
For a deeper understanding of ENABLE and its applications, interested parties are encouraged to visit the official Southwest Research Institute web page dedicated to this innovative technology.
Subject of Research: Markerless motion capture for metabolic cost prediction in biomechanics and rehabilitation.
Article Title: Revolutionizing Metabolic Cost Estimation: ENABLE™ Markerless Motion Capture and Machine Learning Transform Rehabilitation and Performance Assessment.
News Publication Date: June 23, 2026.
Image Credits: Southwest Research Institute
Keywords
Markerless motion capture, ENABLE™, metabolic cost prediction, musculoskeletal modeling, machine learning, biomechanics, rehabilitation, prosthetics, gait analysis, energy expenditure, computer vision, OpenSim models.
