SAE-based classification of school-aged children with autism spectrum disorders using functional magnetic resonance imaging

摘要: This paper employs a novel-deep learning method and brain frequencies to discriminate school-aged children with autism spectrum disorders (ASD) from typically developing (TD) functional magnetic resonance imaging (fMRI) data of 84 subjects the ABIDE (Autism Brain Imaging Data Exchange) database. Firstly, fMRI were preprocessed, then each subject’s dataset was decomposed into 30 independent components (IC). Secondly, some key ICs selected inputted stacked autoencoder (SAE). The SAE adopted for features subtraction dimensionality reduction. Finally, softmax classifier used ASD TD children. average accuracy work as high 87.21% (average sensitivity = 92.86%, specificity 84.32%). results classification demonstrated that proposed may have potential automatically Attempts use deep learning-based algorithms should likely be step forward in auxiliary clinical utility.