Maximum-Entropy-Rate Selection of Features for Classifying Changes in Knee and Ankle Dynamics During Running

This paper investigates deteriorations in knee and ankle dynamics during running. Changes in lower limb accelerations are analyzed by a wearable musculoskeletal monitoring system. The system employs a machine-learning technique to classify joint stiffness. A maximum-entropy-rate method is developed to select the most relevant features. Experimental results demonstrate that distance travelled and energy expended can be estimated from observed changes in knee and ankle motions during 5-km runs.

I. Introduction

Muscle fatigue is commonly defined as a reduced ability to maintain force or power output during prolonged physical activity. Disturbances in the concentrations of muscle lactate, hydrogen, potassium and calcium ions are linked with fatigue [1]. Fatigue is also associated with a reduction in the metabolic uptake of oxygen by muscle cells during prolonged exercise. Fatigue monitoring is important for three reasons. In fatigue management, sportspeople may need to be rested and revitalized to maximize performance. Second, to aid injury prevention as tendon and muscle functions become impaired with overuse. Apart from subjective feelings of soreness and tiredness, continuing sporting activities during fatigue can adversely affect lower limb abilities to absorb shock and to maintain joint stability. Compromised joint stability can in turn increase the likelihood of strains, sprains, stress fractures and falls. Third, to diagnose post-injury rehabilitation. Objective indicators of joint stability can provide information about the efficacy of support measures (e.g., taping, strengthening, conditioning) and for monitoring recovery.