Evaluation of Autonomy Walk by Dynamic Foot Pressure Analysis

This paper describes a autonomy walk evaluation method based on fuzzy logic for rehabilitation field. The method acquires dynamic foot pressure distribution change while walking by mat type load distribution sensor. To evaluate the walk, four gait features such as gait speed, time of double support, gait balance and total time of stance phase and are extracted from the foot pressure data. From these features, the method derives eight fuzzy if-then rules and fuzzy membership functions. These functions are learned by the foot pressure of commons and patients. A gait level index is estimated by fuzzy degree. In experiment, we employed ten patients and ninety commons, and took their pressure distribution data. To evaluate our method, we classified them to Patient or Common by their gait level indexes. We compared our method with support vector machines. The proposed method obtained higher classification rate.