Handgrip Strength Evaluation Using Neuro Fuzzy Approach

Abstract

Handgrip assessment is a useful method to monitor patient rehabilitation. The neurofuzzy analysis provides system identification and interpretability of fuzzy models and learning capability of neural networks. The purpose of this study is to collect handgrip strength of patients and distinguish them from the normal persons. Multilevel Perception neural network utilizes the back-propagation learning algorithm is suitable to discover relationships and patterns in the dataset. When the parameters are well tuned, the expert rules in the training data are captured and stored as expert weights of the neural network. The expert rules define the membership function for the fuzzy system. The fuzzy model based on the membership function, fed in by the neural network will intelligently classify the data. The results indicate that the classification accuracy of normal and pathological patients are 90% and 75% respectively. Moreover, this research demonstrates the feasibility of a novel handgrip design because the force measurements variance of the conventional LIDO machine and our designed handgrip is only 0.169.