Convolutional Neural Network based Autism Classification
F. Catherine Tamilarasi, Jayanthi Venkatraman Shanmugam
Abstract
Autism spectrum disorder is a disease in brain development that affects the way a person socializes with others and cause problems in interaction and communication. The limited and repetitive patterns of behavior are also observed in this disorder. The word Spectrum refers to a broad range of symptoms and difficulties. Deep neural networks achieve good performance in several applications. In this method, a convolution neural network is proposed to instinctually identify young children with possibilities of ASD in premature age. Utilizing the Convolution Neural Network (CNN) ResNet-50 architecture, the images of Autism spectrum disorder is arranged from normal controls. The method exhibit that the effective results for both sensitivity and specificity are acquired.