Abstract
Osteoarthritis, stroke, spinal cord injury, traumatic brain injury, and amputation affect roughly 19% of the U.S.
adult population, with osteoarthritis and stroke being leading causes of serious long-term disability in adults
worldwide. Along with other conditions such as cerebral palsy, Parkinson's disease, and orthopedic cancer,
these conditions often significantly impair movement, resulting in substantial societal costs, an increased risk of
other serious health conditions (e.g., heart disease and diabetes), a reduction or even loss of independence,
and a decreased quality of life. Despite the significance of the problem and the uniqueness of each patient,
treatment design for movement impairments has not progressed substantially beyond off-the-shelf
interventions selected based on subjective clinical judgment. If affected individuals are to recover the most
function possible, a paradigm shift is needed toward personalized interventions designed using objective
evidence-based methods.
This project seeks to develop innovative software technology that will allow engineers working in
collaboration with clinicians to design effective personalized interventions for movement impairments using
objective physics-based computer models. The software technology will employ the same computer modeling
and simulation methods that have revolutionized the design of airplanes and automobiles over the past 25
years. The proposed software will create a virtual representation of the patient and then apply virtual
treatments to the virtual patient to identify the treatment design that is most likely to maximize recovery of lost
function. Virtual patient models will obey laws of physics and principles of physiology to reflect how the patient
moves before treatment and predict how the patient will move after treatment. To enable fast and easy
construction of patient models and optimization of patient functional outcomes, the software technology will be
incorporated into the NIH-funded OpenSim software for modeling and simulation of human movement.
To support development and adoption of the proposed software, the project will also use the software to
design personalized interventions for three individuals post-stroke with impaired, asymmetric walking function.
The research team will organize a three-year “Stroke Grand Challenge Competition,” held each year at the
same professional conference, to engage the research community in model-based personalized treatment
design. An extensive human movement data set will be collected from each subject to be used for constructing
a virtual model of the subject. Competing research teams will use the software and the subject's virtual model
to design personalized treatments that improve the subject's walking symmetry. In addition, the research team
will use the new software to develop its own personalized intervention designs for the same subjects. Any
clinically promising interventions identified by either competition participants or the research team will be
implemented on the same subjects in a follow-on project to evaluate their efficacy.
