Towards Better Gait Predictions: Sensor-Based Detection of Flexion and Extension of Human Lower Limb Joints During Walking

Lower limb joint kinetics and dynamics are necessary for the efficiency of walking. As the gait behavior changes, lower body joints like lumbar, knee and ankle undergo constant changes in accordance with the walking environment, physiological and structural changes. These changes can be monitored by the external system and provide a balance mechanism to accomplish the gait pattern. In this study we used low-cost mobile phone accelerometer sensors to extract the data from lower body joints like lumbar, knee and ankle to analyze the kinetic and dynamic behavior in terms of joint velocities, angles during flexion and extension across people with different weight groups. The extracted data has also been classified using different machine learning algorithms to understand the data inhomogeneities and develop a model for better gait prediction. Joint angles varied across the people with different weight groups, where ankle and hip showed modified angular velocities with respect to flexion and extension. Compared to the joint angle data, angular velocity data was better able to discriminate between the gait patterns of the healthy people and provide interventions during gait disorders especially hip related disorders.