DEEP LEARNING BASED AUTISM DISEASE DETECTION

Autism spectrum disorder (ASD) is a neuro developmental condition that affects individuals’ social abilities. Most current approaches rely on structural and resting-state functional magnetic resonance imaging (fMRI) for ASD detection using small datasets, achieving high accuracy but limited generalizability. To improve the effectiveness of automated autism diagnosis with limited data, this study introduces an ASD detection model based on deep learning (DL). A convolutional neural network classifier is employed to address the classification challenge. Simulation results indicate that the proposed model outperforms state-of-the-art methods in terms of accuracy. Specifically, the model explores the use of anatomical and functional connectivity indicators to determine ASD presence, achieving a classification accuracy of 93.41% and a mean absolute error (MAE) of 0.29 in data localization.