Support vector machine (SVM)‐based multivariate pattern analysis (MVPA) has delivered promising performance in decoding specific task states based on functional magnetic resonance imaging (fMRI) of the human brain. Conventionally, the SVM‐MVPA requires careful feature selection/extraction according to expert knowledge. In this study, we propose a deep neural network (DNN) for directly decoding multiple brain task states from fMRI signals of the brain without any burden for feature handcrafts. We trained and tested the DNN classifier using task fMRI data from the Human Connectome Project's S1200 dataset (N = 1,034). In tests to verify its performance, the proposed classification method identified seven tasks with an average accuracy of 93.7%. We also showed the general applicability of the DNN for transfer learning to small datasets (N = 43), a situation encountered in typical neuroscience research. The proposed method achieved an average accuracy of 89.0 and 94.7% on a working memory task and a motor classification task, respectively, higher than the accuracy of 69.2 and 68.6% obtained by the SVM‐MVPA. A network visualization analysis showed that the DNN automatically detected features from areas of the brain related to each task. Without incurring the burden of handcrafting the features, the proposed deep decoding method can classify brain task states highly accurately, and is a powerful tool for fMRI researchers.

Autism spectrum disorder (ASD) is a general neurodevelopmental disorder associated with altered brain connectivity. However, most connectivity analyses in ASD focus on static functional connectivity, largely neglecting brain activity dynamics that have been reported to provide deeper insight into the underlying mechanisms of brain functions. Therefore, we anticipate that the use of dynamic functional connectivity (DFC) with interaction of clustering measures could help characterize ASD severity and reveal more information. In this study, we applied the sliding‐window and k‐means clustering methods to perform DFC and clustering analyses in ASD and typically developing (TD) groups. Data from 62 ASD and 63 TD children were acquired from the open‐access data set Autism Brain Imaging Data Exchange. Our findings revealed higher DFC variability between the posterior cingulate gyrus (PCC) and middle temporal pole (TPOmid) in subjects with ASD. The connection between the PCC and pars opercularis of inferior frontal gyrus (IFGoper) also presented greater variability in ASD, with the increase depending on ASD symptom severity. Furthermore, clustering analysis showed higher averaged dwell time and probability of transition for globally hyper‐connected state in the ASD group, which could be related to connection variability between the PCC and IFGoper. Our results demonstrate that both the PCC and IFGoper play crucial roles in characterizing symptom severity and state configuration in ASD, and brain connectivity dynamics may serve as potential indicators of ASD in future studies. Autism Res 2020, 13: 230–243. © 2019 International Society for Autism Research, Wiley Periodicals, Inc. Lay Summary Dynamic functional connectivity (DFC) refers to functional connectivity that changes over a short time. This study found that DFC instability between the posterior cingulate gyrus and pars opercularis of inferior frontal gyrus is associated with abnormal brain pattern configurations and dysfunction of social cognitive processes in autism spectrum disorder (ASD). These findings could contribute to a deeper understanding of the neural mechanisms of ASD and help characterize ASD severity.

No abstract

Purpose Previous studies have focused on global cerebral alterations observed in cirrhosis. However, little was known about the specific abnormalities of vision-related brain regions in cirrhotic patients. In this study, we sought to explore neurological alterations of vision-related regions by measuring brain resting-state network connectivity, based on the structural investigation in cirrhotic patients without clinical sign of hepatic encephalopathy (HE). Methods Structural and functional magnetic resonance image (MRI) data were collected from 20 hepatitis B virus (HBV)-related cirrhotic patients without clinical sign of HE and from 20 healthy controls (HC). Voxel-based morphometric (VBM) analysis and brain functional network analysis were performed to detect abnormalities in cerebral structure and function. Results Cirrhotic patients showed regions with the most significant gray matter reduction primarily in vision-related brain regions, including the bilateral lingual gyri, left putamen, right fusiform gyrus, and right calcarine gyrus, and other significant gray matter reductions were distributed in bilateral hippocampus. Based on structural investigation focused on vision-related regions, brain functional network analysis revealed decreased functional connectivity between brain functional networks within vision-related regions (primary visual network (PVN), higher visual network (HVN), visuospatial network (VSN)) in the patient group compared with HC group. Conclusion These results indicate that structural and functional impairment were evident in the vision-related brain regions in cirrhotic patients without clinical sign of hepatic encephalopathy. The physiopathology and clinical relevance of these changes could not be ascertained from the present study, which provided a basis for further evolution of the disease.