Neuroimaging Underpinnings Of Auditory Hallucinations In Schizophrenia

BackgroundOne of the major obstacle for the identification of neurobiological underpinnings of schizophrenia(SZ)is the overwhelming clinical heterogeneity of this psychiatric disorder.Dividing SZ into symptom-specific subtypes may reduce heterogeneity to some extent.Auditory hallucinations are one of the cardinal symptoms of SZ,as a main diagnostic criterion in SZ.Elucidation of neural mechanisms underlying auditory hallucinations in SZ is of great clinical significance for treatment of this debilitating mental disease.(1)Auditory verbal hallucinations(AVHs)in SZ involve the abnormality of many brain regions,but the pattern of cerebral blood flow(CBF)and its connectivity are still not well understood.(2)Impaired rich club organization exists in the structural network of SZ,leading to disrupted functional brain dynamics.As a core symptom,altered rich club organization in AVHs has been reported.(3)Altered resting state networks(RSNs)have been proposed to underpin neural mechanisms of AVHs in SZ.However,AVHs-specific patterns within these brain networks remain unclear in SZ based on the direct evidence in vivo.(4)Cerebral dysfunction may represent pathophysiological underpinnings behind AVHs in SZ.However,regional and network functional deficits for AVHs in SZ remain to be identified.(5)Evidence suggests the disturbed interactions among auditory and language processing-related brain regions may be crucially implicated in the pathophysiology of AVHs in SZ.However,information flow within these brain networks remains unclear.Consequently,the current study explored the neuroimaging underpinnings of AVHs in SZ by means of CBF and its complex networks,rich club organization,RSNs,regional and network function,and effective connectivity using functional magnetic resonance imaging(f MRI),thereby shedding light on the neural mechanism behind it.MethodsUsing resting-state pulsed arterial spin labeling(p ASL),diffusion tensor imaging(DTI),blood oxygen level-dependent-functional magnetic resonance imaging(BOLD-f MRI)and high-resolution structural imaging,SZ patients with and without AVHs,as well as healthy controls(HCs)were included in this study.Both first-episode and non-first-episode SZ patients were enrolled in(1)(n = 25 in each group),but only medication-na?ve first-episode patients were included in(2)(n = 15 in AVHs group,n = 27 in Non-AVHs group,and n = 48 in HCs).As well,only first-episode patients were included in(3),(4) and(5)(n = 17 in AVHs group,n= 15 in Non-AVHs group,and n = 19 in HCs group).(1)The CBF values were measured and compared between each group,the CBF connectivity was calculated,and CBF topological properties were further analyzed,including clustering index,characteristic path length,local efficiency,and global efficiency.(2)Measures of rich club organization,connectivity density of rich club,feeder,and local connections were performed in structural connectivity of the cerebral cortex.(3)Independent component analysis(ICA)was used to investigate RSNs.Dual regression was implemented to obtain between-subject analysis.We then explored regional brain function within RSNs using amplitude of low-frequency fluctuation(ALFF).Finally,we performed correlation and ROC analyses.(4)The ALFF and regional homogeneity(Re Ho)were compared among these subjects.Areas with both ALFF and Re Ho alterations were used as seeds in functional connectivity(FC)analysis.Then we performed correlation analysis between image measures and symptoms and receiver operating characteristic(ROC)analysis.(5)Stochastic dynamic causal modeling(s DCM)was used to quantify directed connections among distinct brain regions,including the left dorsolateral prefrontal cortex(inner speech monitoring),auditory cortex(auditory processing),hippocampus(memory retrieval),thalamus(information filtering),and Broca's area(language producing).Results(1)CBF and its connectivity features of AVHs in SZ: dysfunctional speech monitoring,imagery,and productionAVHs patients exhibited decreased CBF in the bilateral superior and middle frontal gyri and postcentral gyri,and right supplementary motor area compared with SZ patients without AVHs.SZ patients without AVHs showed reduced CBF in the left middle frontal gyrus relative to HCs.Moreover,AVHs groups showed distinct connectivity pattern,an intermediate level between HCs and patients without AVHs in the global efficiency.(2)Disrupted structural network of the brain in first-episode schizophrenia with auditory verbal hallucinationsRich club phenomenon was found in the structural networks of SZ patients,with a significant reduction of rich club organization.However,there was no significant difference of rich club organiztion between patients with and without AVHs.Furthermore,relative to Non-AVHs,AVHs patients showed no significant difference in rich club and feeder density but a significant reduction in local density.(3)Disturbed brain activity in resting state networks of first-episode schizophrenia with auditory verbal hallucinationsICA revealed symptom-specific abnormal disrupted coactivation within the auditory,default mode,executive,motor,and frontoparietal networks,pronouncedly in the auditory cortex,supramarginal gyrus,insula,putamen,dorsolateral prefrontal cortex,angular gyrus,precuneus,and thalamus.ALFF analysis revealed the similar patterns within these RSNs.Furthermore,a positive correlation between the degree of coactivation within the motor network and the severity of AVHs was observed in SZ patients with AVHs.Area under ROC curve was larger than 0.75 for all the RSNs.(4)Putamen-related regional and network functional deficits in first-episode schizophrenia with auditory verbal hallucinationsOne-way analysis of variance showed significant differences of ALFF and Re Ho in the bilateral putamen,thereby being used as seeds.SZ patients with AVHs showed decreased ALFF in the left putamen,increased Re Ho in the right dorsolateral prefrontal cortex,and increased right putamen-seeded FC with the left dorsolateral prefrontal cortex and Broca's area relative to those without AVHs.Furthermore,the increased strength of the connectivity between the right putamen and left Broca's area correlated with the severity of SZ symptoms.Both patient groups demonstrated hypoconnectivity within frontal/parietal/temporal cortico-striatal-cerebellar networks compared with HCs.(5)Thalamic-auditory cortical-hippocampal dysconnectivity in first-episode schizophrenia patients with auditory verbal hallucinationss DCM revealed symptom-specific abnormal effective connectivity involving the thalamic-auditory cortical-hippocampal circuit in SZ patients with AVHs,with an increased sensitivity of auditory cortex to its thalamic afferents and a decrease in hippocampal sensitivity to auditory inputs.Furthermore,a positive correlation between the strength of the connectivity from Broca's area to the auditory cortex and the severity of AVHs was observed in SZ patients with AVHs.Conclusion(1)The current study demonstrates aberrant CBF in the brain regions associated with inner speech monitoring and language processing in SZ patients with AVHs.The complex network measures showed by CBF-derived functional connectivity indicate dysconnectivity between different functional units within the network of AVHs in SZ.These findings might shed light on the neural underpinnings behind AVHs in this disease at the level of CBF and its connectivity.(2)From the perspective of rich club organization of structural network,as a main symptom,AVHs in SZ does not reflect significant difference with patients without AVHs.Moreover,AVHs have association with connections non-rich club regions themselves,but not among rich club regions or rich club and non-rich club regions.(3)These findings indicated dysfunctional brain regions involving auditory processing,language producing and monitoring,and sensory information filtering in AVHs in SZ,predisposing to false perceptual inference.This study may provide a novel insight into strategy for diagnosis and therapy of AVHs in SZ based on imbalance among those RSNs.(4)AVHs in SZ may be caused by abnormal regional function in the putamen and prefrontal cortex,as well as hyperconnectivity between them.The putamen-related regional and network functional deficits may reflect imbalance in neuromodulation of AVHs in SZ.Furthermore,dysconnectivity within cortico-striatal-cerebellar networks might subserve the pathogenesis of SZ.(5)These findings for the first time indicate augmented excitatory afferents from the thalamus to the auditory cortex in SZ patients with AVHs,resulting in auditory perception without external auditory stimuli.Our results may provide insights into the neural mechanisms underlying AVHs in SZ.Thalamic-auditory cortical-hippocampal dysconnectivity may also serve as a potential diagnostic biomarker and therapeutic target of AVHs in SZ based on the direct evidence in vivo we found.These findings indicate that language-related memory retrieval helps to provide the source of inner speech,and active speech imagery,initiation,and production without appropriate monitoring yields excessive inner speech,inducing overmuch auditory input due to deficit information filtering.Therefore,SZ patients with AHVs suffer from disrupted structure of the large-scale brain network.Regional dysfunction and abnormal interaction of brain regions involving auditory sense,language,memory and sensory information filtering might underly the neuroimaging underpinngings of AVHs in SZ,predisposing patients with SZ toward having auditory hallucinations in the absence of external auditory stimuli.