默认模式网络
神经科学
自闭症谱系障碍
隐马尔可夫模型
神经影像学
心理学
自闭症
静息状态功能磁共振成像
功能磁共振成像
网络动力学
动态功能连接
大脑活动与冥想
计算机科学
人工智能
发展心理学
脑电图
离散数学
数学
作者
Pingting Lin,Shiyi Zang,Y. Bai,Haixian Wang
标识
DOI:10.3389/fnhum.2022.774921
摘要
Autism spectrum disorder (ASD) is a group of complex neurodevelopment disorders characterized by altered brain connectivity. However, the majority of neuroimaging studies for ASD focus on the static pattern of brain function and largely neglect brain activity dynamics, which might provide deeper insight into the underlying mechanism of brain functions for ASD. Therefore, we proposed a framework with Hidden Markov Model (HMM) analysis for resting-state functional MRI (fMRI) from a large multicenter dataset of 507 male subjects. Specifically, the 507 subjects included 209 subjects with ASD and 298 well-matched health controls across 14 sites from the Autism Brain Imaging Data Exchange (ABIDE). Based on the HMM, we can identify the recurring brain function networks over time across ASD and healthy controls (HCs). Then we assessed the dynamical configuration of the whole-brain networks and further analyzed the community structure of transitions across the brain states. Based on the 19 HMM states, we found that the global temporal statistics of the specific HMM states (including fractional occupancies and lifetimes) were significantly altered in ASD compared to HCs. These specific HMM states were characterized by the activation pattern of default mode network (DMN), sensory processing networks [including visual network, auditory network, and sensory and motor network (SMN)]. Meanwhile, we also find that the specific modules of transitions between states were closely related to ASD. Our findings indicate the temporal reconfiguration of the brain network in ASD and provide novel insights into the dynamics of the whole-brain networks for ASD.
科研通智能强力驱动
Strongly Powered by AbleSci AI