A portable measurement system for the evaluation of human gait

A tool has been developed which can be used to evaluate human gait in a more detailed manner. Its purpose is to record data from an individual during gait then categorize and analyze the intrinsic phases with neuro-fuzzy techniques. The system is simple to use, adaptive, highly mobile, and does not require calibration. The hardware consists of four accelerometers and four force sensitive resistors to record data during walking which is then prepared and collected by a digital device and PDA computer. The gait data is passed into an intelligent fuzzy inference system managed by custom defined fuzzy rules to be classified into four stance phases (heel strike, flat foot, heel lift, toe push-off), and three swing phases (initial flexion, terminal flexion, and terminal extension). After the fuzzy system was trained using data from five healthy subjects, the system's representative gait classification root mean squared error dropped from 0.2975 to 0.1200, showing a much improved ability to categorize human gait, despite its varied nature. The system represents a robust tool, which can be used in a clinical environment for the analysis of human gait in rehabilitative applications such as rule based control generation for functional electrical stimulation, and gait quality analysis.