Autism Spectrum Disorder DetectionUsingMachine Learning

Convolutional Neural Networks (CNNs) have made significant strides in various domains such as image classifica- tion, speech recognition, automotive software engineering and neuroscience. In this study we leverage CNNs for the automated detection of Autism Spectrum Disorder using brain imaging data,specificallywe utilize resting-state functional Magnetic Resonance Imaging (f-MRI) data from the Autism Brain Imaging Exchange (ABIDE) dataset. To identify ASD patients, our approach relies on analyzing patterns of functional connectivity to classify ASD and control subjects. Experimental results demonstrate that our proposed model achieves an accuracy of 96.40% on the ABIDEI dataset using the CC400 functional parcellation atlas of thebrain. Notably our CNN model exhibits computational efficiency by employing fewer parameters comparedtostate-of-the-art techniques. Our findings suggest that the developed model shows promise for prescreeningasdpatients and warrants further evaluation with additional datasets.