شناسايي اختلال طيف اوتيسم مبتني بر بررسي توام داده‌هاي ساختاري و عملكردي با استفاده از شبكه‌هاي عصبي گرافي

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by difficulties in social interaction an‎d communication, as well as the presence of repetitive behaviors. Conventional methods for diagnosing autism are primarily based on clinical interviews an‎d direct observation of patientsʹ behavior. These procedures are time-consuming, highly dependent on the expertise of the eva‎luator, an‎d prone to potential errors. Early diagnosis of this disorder can play a significant role in improving the quality of life of affected individuals an‎d their families. With advances in neuroimaging techniques, particularly magnetic resonance imaging (MRI), it has become possible to investigate the structure an‎d function of the brain in a noninvasive manner. One of the emerging approaches in neuroscience research is the integration of structural an‎d functional information from different brain regions in order to achieve a better understan‎ding of brain networks an‎d to identify neurological disorders such as autism. In this context, the spatial an‎d temporal relationships among brain regions can be modeled as graphs, enabling a more precise analysis of brain connectivity patterns. In this study, spatial-temporal features were extracted from brain graphs constructed based on structural an‎d functional data - namely diffusion tensor imaging (DTI) an‎d functional magnetic resonance imaging (fMRI) - using the Graph WaveNet model. Subsequently, a support vector machine (SVM) classifier was employed to distinguish between healthy individuals an‎d those with autism spectrum disorder. To eva‎luate the performance of the proposed model, five-fold cross-validation an‎d voting strategies were applied. The results demonstrated that the proposed model achieved an average accuracy of 80%, with a maximum accuracy of 88% in some folds, indicating a promising performance in differentiating healthy individuals from those with ASD.

زينب مالكي , فرزانه شايق بروجني

جلال ذهبي , نيلوفر احمدي پور