Pooya Mazloomi - MSc Student of Industrial Engineering, School of Industrial and Systems Engineering, Tarbiat Modares University (TMU)
Niloufar Naddafi - MSc Student of Industrial Engineering, School of Industrial and Systems Engineering, Tarbiat Modares University (TMU)
Toktam Khatibi - Associate Professor, School of Industrial and Systems Engineering, Tarbiat Modares University (TMU)

Autism spectrum disorders (ASD) are serious cognitive disorders that affect people's social communication and daily routine. In recent years, pattern recognition methods in neuroimaging data have become more important than before, and this approach can help doctors diagnose cognitive disorders. Magnetic resonance imaging (MRI) is one of the most widely used imaging methods to study the neural basis of cognitive disorders, which is widely used for accurate diagnosis of ASD. However, MRI analysis for the diagnosis of ASD is a specialized and time-consuming task. Therefore, some previous studies have focused on proposing automated methods for MRI processing to diagnose disorders. The main purpose of this study is to review previous studies of ASD diagnosis based on automated methods with the aim of highlighting the use of smart technology in the autism diagnosis process and investigating its impact on the autism diagnosis process. This review considers previous papers in terms of datasets, preprocessing tasks, data representation methods, models and classifiers. The use of automatic methods in the analysis of brain images improves the accuracy and speed of ASD diagnosis and can help doctors make a better diagnosis.

Diagnostic Process Enhancement, Medical image analysis, Early Detection, Neural Networks, Al in Healthcare