Real-time Gait Pattern Classification Using Artificial Neural Networks

The expression of human gait associated with neurological disorders is difficult to describe and is characterized by fluctuating predominance in the presence of complex movement patterns. The analysis of human gait patterns can provide significant information related to the physical and neurological functions of individuals, and may contribute to the diagnosis of human motor disorders in pathological conditions.The present study seeks to determine the classification capacity of different types of simulated abnormal gait patterns by recording the accelerations of the center of mass, the extraction of characteristics in the time and frequency domain and the classification based on the use of artificial neural networks in real time.