ADJUSTING OPTIMAL HEIGHT OF WALKER HANDLES USING EXPERIMENTAL COGNITIVE DATA AND ADAPTIVE NEURO-FUZZY CLASSIFICATION

Abstract

Background: Walker is one of the most important devices to help people who need assistance to maintain balance or stability while walking. A walker is often used by people who deal with reduced muscle strength of the lower limbs or reduced range of motion in hip, knee or ankle. One of the things that should be considered when using walker is its handle’s optimal height. This height could be different according to the physical parameters of each person’s height, weight, gender and age and can significantly affect the level of assistance and ergonomic aspects of the device. Methods: In this article, the basis for manufacturing an intelligent walker is presented by proposing an algorithm for determining the optimal height of the handle according to the users’ characteristics. To achieve this, an experimental test was first performed on 47 subjects. After performing this experimental test, the information about each person, as well as the level of satisfaction declared by the user for the specific height was recorded. Then an intelligent algorithm was generated by an adaptive neuro-fuzzy classifier and trained accordingly. The input of the classifier is the user characteristics and the output is the optimal height. Results: The algorithm predicted the optimal height with 87% accuracy. The result of the test was to obtain an optimal height for the subjects tested. The results obtained from the classifier show that optimal height depends on factors such as gender, age, height and weight. Conclusion: In this method, in order to obtain the optimal height for each person, one of the main parameters was the degree of satisfaction of the individual, which was not taken into account in the previous studies in this field. In this intelligent algorithm, by entering information of an individual, his/her optimal height can be calculated. The proposed method can be utilized in improving intelligent walkers.