Fuzzy Classifier based on Muscle Fatigue Parameters

This paper presents the development of a decision aid tool based on a fuzzy classifier. The goal was to obtain a system that could support a physician who have to make decisions about how to deal with the progression of the disease of a child affected by Duchenne Muscular Dystrophy. First, we used an outranking multicriteria method to select among the possible parameters of muscle fatigue evaluation the subset that provide more reliable information, than we used the selected parameters as membership functions for the fuzzy classifier. The output of the fuzzy classifier consisted of three classes: 1-"close to normal results", 2-"results compatible with moderate pathological conditions", and 3-"results congruent with severe pathological conditions". A first test of the classifier was performed using the data of the twenty examinations of six children and it provided good results. We believe that these results are relevant to the clinical applications and they can be easily extended to different pathologies.