Quantum convolution searched binary neural networks based autism spectrum disorder detection using MRI images in cloud computing.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disease that causes discrepancies in social interaction and behavioral changes. The developments of neuroimaging techniques, like Magnetic Resonance Imaging (MRI) is employed to detect brain abnormalities. Due to the heterogeneity of disease severity and symptoms, the detection of ASD is difficult. To solve such complexity, a novel model named Fractional Painting Training Based Optimization trained Quantum Convolution Searched Binary Neural Network (FPTO_QCSBNN) is proposed for ASD detection in cloud. A cloud-based detection system offers the analysis and storage of large-scale neuroimages. Moreover, it provides faster diagnosis with scalable storage. Initially, the cloud system is simulated, and pre-processing is done using Mid-Point filter and Region of Interest (ROI) extraction. Image enhancement is done by gamma correction method, and pivotal region is extracted using functional connectivity. The optimal grid selection in pivotal region extraction is done using FPTO, and features are extracted from enhanced image. Using features and pivotal region extracted image, QCSBNN detects ASD, and it is trained by FPTO. Furthermore, developed FPTO_QCSBNN attains the accuracy, True Positive Rate (TPR), and True Negative Rate (TNR) of 91.37%, 91.32%, and 91.89%.