Assessment of gait patterns using neural networks.

A new approach for the assessment of gait patterns is presented. The use of neural network techniques for decision making in gait analysis is for some purposes more effective than biomechanical methods or conventional statistics. To demonstrate this, a neural network was trained to distinguish 'healthy' from 'pathological' gait. The algorithm presented here can be used for several purposes because it learns from examples of diagnosed gait patterns without having any built-in model of gait.