Gait asymmetry detection in older adults using a light ear-worn sensor

Abstract

Measuring gait asymmetry is an important feature when characterizing functional imbalance between limbs. This could be due to pathologies, such as osteoarthritis, stroke, or associated with the effects of surgeries such as hip arthroplasty. Generally, the study of asymmetry or imbalance has required the use of a gait lab or force plates, which could be expensive and difficult to use in home settings. This work validates the use of a light weight ear sensor (7.4 g) with an instrumented treadmill for 64 subjects (age (60.04 (15.36)) including healthy subjects (14) as well as subjects who had been treated for hip (17), knee-replacement surgery (21) and knee osteoarthritis (12). Subjects performed treadmill walking at several speeds on flat surfaces, inclines and declines. Accelerometer data from the ear sensor were segmented into consecutive steps and temporal features were extracted. The measures of gait cycle time and step-period asymmetry obtained from the ear sensor matched well those of the treadmill for flat surfaces, inclines and declines. The key implication of the study is that the proposed method could replace expensive equipment for monitoring temporal gait features in clinics as well as free-living environments, which is important for monitoring rehabilitation after surgery and the progress of diseases affecting limb imbalance.