| Objective: Sensorineural hearing loss(SNHL)is the most common congenital sensory deficit of newborns.Without hearing rehabilitation,hearing loss can cause detrimental effects on speech,language,and cognitive outcomes in children.Cochlear implantation(CI)is an important procedure for the restoration of hearing for children with severeto-profound SNHL.Children with congenital SNHL can benefit most from CI before the age of 3.5.In humans,this period is the developmental critical period,in which the cerebral cortex has maximum plasticity.Multimodal magnetic resonance analysis,including blood oxygen level dependent(BOLD)functional magnetic resonance imaging(f MRI)and diffusion tensor imaging(DTI),provides important imaging information for understanding the structural and functional changes of central nervous system in disease states.In recent years,multimodal magnetic resonance analysis has been widely used in the neuropathological mechanism and clinical research of patients with hearing loss.Some studies have found that changes in the structure and function of some brain regions are related to the clinical manifestations or prognosis of hearing loss.However,the participants of most of these studies were adults,whose cortical plasticity is significantly reduced compared with children during the developmental critical period.In addition,the effects of hearing loss are not limited to only several systems and resting-state networks but may result in changes in the whole brain network.Previous f MRI and DTI studies mostly selected several brain regions as areas of interest but paid less attention to the coordination of the brain as a whole.In the present study,we analyze the alteration in resting-state functional brain network,dynamic functional brain network and white matter structural network of SNHL infants within the developmental critical period using BOLD and DTI sequence.We aim to improve the understanding of the effects of auditory loss on function and white matter structure development during this special period.Methods: Children in the SNHL group were under three years old with profound bilateral congenital SNHL.Children in the control group were normal-hearing within the same age range.The BOLD and DTI images were collected on a Siemens Verio Tim 3.0 T MR scanner.The MRI data were analyzed as follow: 1.Resting-state f MRI data: The functional images of 34 SNHL children and 30 normalhearing children were parcellated into 90 brain regions based on the University of North Carolina(UNC)two-year-old infant atlas.Each region was considered as a node of the network and the functional connectivity strength of each two nodes was considered as an edge,resulting in a functional connectivity matrix.Differences in global and nodal topologic organization,hub distribution,and whole-brain functional connectivity were explored using complex network analysis.2.Resting-state f MRI data: Included 32 SNHL children and 24 normal-hearing children.Using independent component analysis(ICA),18 independent components were identified to form five resting-state brain networks(RSN).Each independent component was considered as a node and the functional connectivity between every two independent components was considered as an edge.A sliding windows approach and a k-means algorithm was applied to analyze the dynamic functional network connection states of all participants.In each state,the differences of functional connection strength between independent components and network dynamic characteristics between SNHL children and control children were compared.3.DTI data: Included 34 SNHL children and 30 normal-hearing children.The UNC two-year-old infant atlas was used to definite the nodes of the network.Then,the probabilistic tracking method was used and the connection probability between each two nodes was considered as an edge,resulting in a white matter structural connectivity matrix.Differences in topologic organization,rich-club organization,and structural connection were explored using complex network analysis.Results: 1.Static functional network: There was no significant difference in the global parameters of the functional network between the SNHL group and the control group.The SNHL group showed an increased nodal degree in the left inferior marginal angular gyrus and a decreased nodal degree in the bilateral thalamus(P < 0.05,Bonferroni corrected).We found six hubs in the SNHL group and seven hubs in the control group.The left medial superior frontal gyrus and right inferior temporal gyrus were hubs in both the SNHL group and the control group.The right dorsolateral superior frontal gyrus,bilateral orbital part of the inferior frontal gyrus,and left middle temporal gyrus were hubs only in the SNHL group,while the left medial orbital superior frontal gyrus,right parahippocampal gyrus,bilateral temporal pole of the superior temporal gyrus,and left temporal pole of the middle temporal gyrus were hubs only in the control group.The SNHL group presented a significantly decreased subnetwork component consisted of seven nodes and six edge(P < 0.001,NBS corrected).The involved regions included the bilateral median cingulate and paracingulate gyri,left precuneus,right paracentral lobule,bilateral middle temporal gyrus,and right temporal pole of the middle temporal gyrus.2.Dynamic functional network: We identified three states using k-means algorithm.In State 1,the functional network showed generally weak connectivity between and within RSNs.In State 2,independent components of visual network(VN)had a strong positive and negative intra-network connectivity,as well as a strong negative inter-network connectivity with sensorimotor network(SMN)and executive control network(ECN).State 3 showed more positive connectivity between default-mode network(DMN)and SMN,between ECN and VN,and within SMN.Furthermore,compared with the control group,the SNHL group showed decreased connectivity strength between auditory network(AUD)and SMN(P < 0.05,FDR correlated).There was no significant difference in variance of global efficiency and local efficiency between the SNHL group and the control group.3.White matter structural network: There was no significant difference in the global and nodal parameters of the structural network between the SNHL group and the control group.Ten rich nodes were identified and the SNHL group showed significantly higher strength of the local connections compared with the control group(P < 0.05,Bonferroni corrected).Compared with the control group,the SNHL group presented a significantly increased subnetwork component and a significantly decreased subnetwork component(P < 0.01,NBS corrected).The increased subnetwork was a right caudate nucleus centered component consisting of three edges connecting the right caudate nucleus to three different regions,including right precentral gyrus,right medial orbital part of superior frontal gyrus and right globus pallidus.The decreased subnetwork contained nine nodes and eight edges.Right precuneus was a center of this component.Conclusion: Static and dynamic functional network analysis showed that the alterations of regions in hub distribution and functional connectivity of SNHL children were in sensory areas and higher-order cognitive areas,indicating that the brain functional network has undergone adaptive changes in the first few years of hearing deprivation.Meanwhile,the global topological parameters of the static functional network did not change,suggesting that the brain functional network of SNHL children still maintained the ability to transmit information efficiently.The variance of global efficiency and local efficiency of dynamic functional network did not change,indicating that the network stability was maintained.White matter structure network analysis showed that the white matter structure network of SNHL children also underwent adaptive changes due to hearing loss.But the alteration of the functional network and structural network is not one-to-one.While the network maintained its ability to transmit information efficiently,there were redundant connections between non-important brain regions,suggesting that the synaptic pruning process in children with SNHL might be delayed.This study used multi-modal magnetic resonance brain network analysis to explore the changes of brain function and white matter structure in children under three years old with profound bilateral congenital SNHL from the perspective of network.This study providing new insights into the influence of hearing loss on brain development during developmental critical period. |
PDF Full Text Request