Detection of Autism in Children by the EEG Behavior Using Hybrid Bat Algorithm-Based ANFIS Classifier

Abstract

Autism spectrum disorder (ASD) is basically a varied neurodevelopmental disorder which affects the developmental curve in numerous behavioral domains. This includes the impairments of social communication, cognitive and also language abilities. Autism has an extensive gamut of severity and symptoms that are often utilized to categorize the ASD. Each of the syndromes under ASD is disparate as of the other. For instance, individuals suffering with Asperger syndrome have no considerable hindrance in language development. Kids with ASD might concurrently also encompass other issues, for instance, Tourette’s syndrome epilepsies, attention-deficit hyperactivity disorder, etc. So, there is a need to detect autism in children. ASD can be detected in children by analyzing the signals (EEG). Initially, the EEG signals are attained from the dataset. This signal undergoes preprocessing utilizing Kalman filter. Next, signal decomposition is done by the variable mode decomposition. This is followed by feature extraction. These are classified utilizing Hybrid Bat Algorithm with ANFIS classifier (HBA–ANFIS). The classified output is either a normal signal or an autism signal. To prove the proposed work’s efficiency, disparate performance metrics were determined and also analyzed. The proposed work shows superior performance in respect of all the evaluated metrics.