QUANTIFYING BIOMECHANICAL LOADING ON THE ANKLE TO CLINICALLY ESTIMATE ACHILLES TENDON FORCES DURING HOPPING

Abstract

In athletes, Achilles tendinopathy is an overuse injury resulting from biomechanical overload of the Achilles tendon during sporting tasks, such as hopping. Nearly half of individuals with Achilles tendinopathy will fail conservative measures. Therefore, it is paramount to develop effective screening techniques to clinically identify individuals who are at a high risk for Achilles tendinopathy so that effective injury prevention strategies can be prescribed. Chronically high Achilles tendon forces, particularly during eccentric loading, may result in degradation of the tendon matrix. These variables can be accurately measured using three dimensional motion analysis instruments (gold standard), however, this technology is expensive and not readily available to the majority of clinicians. Therefore, the purpose of this study was to develop clinically feasible algorithms to estimate eccentric ankle power and peak Achilles tendon force during hopping using a high-speed video camera. 18 healthy active males who regularly participated in running and jumping activities (Tegner Activity scale score ≥5/10) were recruited. Participants performed 14 stationary single-leg hops paced at 132 beats per minute. Hopping trials were simultaneously recorded using a 2D video camera and 3D motion capture equipment. Peak Achilles tendon force and eccentric ankle power were estimated using 2D kinematic inputs for each hop. These 2D estimates were compared to the values measured by 3D motion capture software (gold standard) using a stepwise multiple linear regression. Eccentric ankle power was predicted using 2D variables within 32.3% of the 3D gold standard (SEE=0.678). Peak Achilles tendon force was predicted using 2D variables within 37.5% of the 3D gold standard (SEE=813.5). Differences in coordinate system origins, parallax when viewing the 2D video, and the small sample size were all possible contributors of the percent errors found when comparing 3D and 2D motion capture. Accuracy is expected to improve when the full sample (n=40) is collected. Future studies will seek to validate these algorithms in a symptomatic population with Achilles tendinopathy, determine the optimal frame rate for 2D motion capture, and develop algorithms for similar tasks in a field setting. As the accuracy of the 2D algorithms outlined in this paper increases, 2D motion capture could not only be used to quantify loading during biomechanical movements, but could be integrated into popular therapy techniques.