Autism Disorder Diagnosis From EEG Signals based on EXtended Neuro-Fuzzy-Fractal model

Today, one of the most dangerous disorders that threatens children is autism which is a neuromuscular disorder. According to recent researches, in every ۸۸ people, one person suffers from autism disorder. The most common symptom of this disorder is the lack of communication skills with other people. Also, motion impairment is another important factor in autism disorder. It is important to identify and recognize autism disorder in early stages to prevent any improve. Therefore, it is necessary to provide an intelligent medical method in this regard due to the fact the symptoms are very dumb. Hence, the research is attempting to provide an intelligent medical diagnostic system to diagnose autism disorder based on valid BIoGS data. The proposed method is that the input of the fuzzy system and the output is determined by the available data. By the opinion of an expert, data is governed by rules. However, because fuzzy logic is lacking in training dataset, the need for an optimized classifier approach is needed, so neuro-fuzzy model used for rule-based training model. Feature extraction is essential part before classification. The inputs data are brain signals or EEGs, so the fractal model can be considered as a feature extraction method because each mapping of a signal with magnification, it is again similar to the same signal as the principle of self-similarity and uniqueness in the fractal. The neuro-fuzzy fractal model used to classify and autism disorder diagnosis in available data. The results represented that the proposed approach obtained ۹۸.۳۹% in terms of accuracy criteria, which has a functional superiority in identifying and detecting autism disorder in comparison with similar previous methods.

کلیدواژه ها:

Autism Disorder ، Intelligent Health Systems ، Neuro-Fuzzy ، Fractal Model