An Efficient Approach for Detection of Autism Spectrum Disorder Using Electroencephalography Signal

Autism spectrum disorder is a brain disorder in children and is detected using magnetic resonance imaging scan and electroencephalography (EEG) signals. This paper presents a system to detect autism using EEG signals. It uses pre-recorded EEG signal, preprocesses it using a digital filter, and extracts features in time and frequency domain using discrete wavelet transform. The features obtained are given to classifiers such as neural networks, support vector machine, K-nearest neighbour (KNN), subspace KNN, and linear discriminant analysis. The results show that the use of subspace KNN provides the best accuracy of 92.8% for time-domain features.