Disrupted controlling mechanism of salience network on default-mode network and central-executive network in schizophrenia

Neuroimaging studies suggest that the human brain consists of intrinsically organized large-scale neural networks. Among those networks, the interplay among default-mode network (DMN), salience network (SN), and central-executive network (CEN)has been widely employed to understand the functional interaction patterns in health and diseases. This triple network model suggests that SN causally controls DMN and CEN in healthy individuals. This interaction is often referred to as the dynamic controlling mechanism of SN. However, such interactions are not well understood in individuals with schizophrenia. In this study, we leveraged resting state functional magnetic resonance imaging (fMRI) data of schizophrenia (n = 67) and healthy controls (n = 81) to evaluate the functional interactions among DMN, SN, and CEN using dynamical causal modeling. In healthy controls, our analyses replicated previous findings that SN regulates DMN and CEN activities (Mann-Whitney U test; p < 10−8). In schizophrenia, however, our analyses revealed the disrupted SN-based controlling mechanism on DMN and CEN (Mann-Whitney U test; p < 10−16). These results indicate that the disrupted controlling mechanism of SN on two other neural networks may be a candidate neuroimaging phenotype in schizophrenia.

The Vital Role of Central Executive Network in Brain Age: Evidence From Machine Learning and Transcriptional Signatures

Recent studies combining neuroimaging with machine learning methods successfully infer an individual’s brain age, and its discrepancy with the chronological age is used to identify age-related diseases. However, which brain networks play decisive roles in brain age prediction and the underlying biological basis of brain age remain unknown. To answer these questions, we estimated an individual’s brain age in the Southwest University Adult Lifespan Dataset (N = 492) from the gray matter volumes (GMV) derived from T1-weighted MRI scans by means of Gaussian process regression. Computational lesion analysis was performed to determine the importance of each brain network in brain age prediction. Then, we identified brain age-related genes by using prior brain-wide gene expression data, followed by gene enrichment analysis using Metascape. As a result, the prediction model successfully inferred an individual’s brain age and the computational lesion prediction results identified the central executive network as a vital network in brain age prediction (Steiger’s Z = 2.114, p = 0.035). In addition, the brain age-related genes were enriched in Gene Ontology (GO) processes/Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways grouped into numbers of clusters, such as regulation of iron transmembrane transport, synaptic signaling, synapse organization, retrograde endocannabinoid signaling (e.g., dopaminergic synapse), behavior (e.g., memory and associative learning), neurotransmitter secretion, and dendrite development. In all, these results reveal that the GMV of the central executive network played a vital role in predicting brain age and bridged the gap between transcriptome and neuroimaging promoting an integrative understanding of the pathophysiology of brain age.

O4.4. REAL TIME FMRI NEUROFEEDBACK TARGETING STG AND DMN REDUCES AUDITORY HALLUCINATIONS

Background Auditory hallucinations (AH) are one of the core symptoms of schizophrenia (SZ) and constitute a significant source of suffering and disability. One third of SZ patients experience pharmacology-resistant AH, so an alternative/complementary treatment strategy is needed to alleviate this debilitating condition. In this study, real-time functional Magnetic Resonance Imaging neurofeedback (rt-fMRI NFB), a non-invasive technique, was used to help 10 SZ patients modulate their brain activity in key brain regions belonging to the network involved in the experience of auditory hallucinations. In two experiments we selected two different brain targets. 1. the superior temporal gyrus (STG) and 2. default mode network (DMN)-central executive network (CEN) connectivity. STG is a key area in the neurophysiology of AH. Hyperactivation of the default mode network (DMN) and of the superior temporal gyrus (STG) in SZ has been shown in imaging studies. Furthermore, several studies point to reduced anticorrelation between the DMN and the central executive network (CEN). Finally, DMN hyperconnectivity has been associated with positive symptoms such as AHs while reduced DMN anticorrelations have been associated with cognitive impairment. Methods In the STG-focused NFB experiment, subjects were trained to upregulate the STG activity while listening to their own voice recording and downregulate it while ignoring a stranger’s voice recording in the course of 21 min NFB session. Visual feedback was provided to subjects at the end of each run from their own STG activity in the form of a thermometer. AH were assessed with auditory hallucination scale pre-NFB and within a week after the NFB session. The DMN-CEN focused NFB experiment was conducted about 1 month later to minimize the carry over effects from the STG-focused NFB and was designed to help SZ patients modulate their DMN and CEN networks. DMN and CEN networks were defined individually for each subject. The goal of the task was to increase CEN-DMN anti-correlations. To achieve that patients were provided with meditation strategies to guide their performance. Feedback was provided in the form of a ball that traveled up if the modulation of DMN-CEN connectivity was successful and traveled down if it was not successful. AH measures were taken before the NFB session and within a week after the session. Results In the STG-focused NFB task, significant STG activation reduction was found in the comparison of pre- relative to post-NFB in the condition of ignoring another person’s voice (p&lt;0.05), FWE-TFCE corrected. AH were also significantly reduced (p&lt;0.01). Importantly, significant correlation was found between reductions in the STG activation and AH reductions (r=.83). In the DMN-CEN focused NFB task, significant increase in the anti-correlations between medial prefrontal cortex (mPFC) and dorsolateral prefrontal cortex (DLPFC) (p&lt;0.05) was observed as well as significant reduction in the mPFC-PCC connectivity (p &lt;0.05), in the pre-post NFB comparisons. AH were significantly reduced in post- relative to pre-NFB comparison (p&lt;0.02). Finally, there was a significant correlation between individual scores in mPFC-STG connectivity and AH reductions. Discussion These the two experiments suggest that targeting both the STG BOLD activation and DMN-CEN connectivity in NFB tasks aimed at AH reduction result both in brain changes and in AH reductions. Together, these results provide strong preliminary support for the NFB use as a means to impact brain function leading to reductions in AH in SZ. Importantly, these results suggest that AH result from brain abnormalities in a network of brain regions and that targeting a brain region belonging to this network will lead to AH symptom reduction.