Resting-state Network Topology Research Articles

Lesions to both somatic and affective pain pathways lead to decreased salience network connectivity.

Human pain is a salient stimulus composed of two main components: a sensory/somatic component, carrying peripheral nociceptive sensation via the spinothalamic tract and brainstem nuclei to the thalamus and then to sensory cortical regions, and an affective (suffering) component, where information from central thalamic nuclei is carried to the anterior insula, dorsal anterior cingulate cortex and other regions. While the sensory component processes information about stimulus location and intensity, the affective component processes information regarding pain-related expectations, motivation to reduce pain and pain unpleasantness. Unlike investigations of acute pain that are based on the introduction of real-time stimulus during brain recordings, chronic pain investigations are usually based on longitudinal and case-control studies, which are limited in their ability to infer the functional network topology of chronic pain. In the current study, we utilized the unique opportunity to target the CNS's pain pathways in two different hierarchical locations to establish causality between pain relief and specific connectivity changes seen within the salience and sensorimotor networks. We examined how lesions to the affective and somatic pain pathways affect resting-state network topology in cancer patients suffering from severe intractable pain. Two procedures have been employed: percutaneous cervical cordotomy (n = 15), hypothesized to disrupt the transmission of the sensory component of pain along the spinothalamic tract, or stereotactic cingulotomy (n = 7), which refers to bilateral intracranial ablation of an area in the dorsal anterior cingulate cortex and is known to ameliorate the affective component of pain. Both procedures led to immediate significant alleviation of experienced pain and decreased functional connectivity within the salience network. However, only the sensory procedure (cordotomy) led to decreased connectivity within the sensorimotor network. Thus, our results support the existence of two converging systems relaying experienced pain, showing that pain-related suffering can be either directly influenced by interfering with the affective pathway or indirectly influenced by interfering with the ascending spinothalamic tract.

Activity flow mapping over probabilistic functional connectivity.

Emerging evidence indicates that activity flow over resting-state network topology allows the prediction of task activations. However, previous studies have mainly adopted static, linear functional connectivity (FC) estimates as activity flow routes. It is unclear whether an intrinsic network topology that captures the dynamic nature of FC can be a better representation of activity flow routes. Moreover, the effects of between- versus within-network connections and tight versus loose (using rest baseline) task contrasts on the prediction of task-evoked activity across brain systems remain largely unknown. In this study, we first propose a probabilistic FC estimation derived from a dynamic framework as a new activity flow route. Subsequently, activity flow mapping was tested using between- and within-network connections separately for each region as well as using a set of tight task contrasts. Our results showed that probabilistic FC routes substantially improved individual-level activity flow prediction. Although it provided better group-level prediction, the multiple regression approach was more dependent on the length of data points at the individual-level prediction. Regardless of FC type, we consistently observed that between-network connections showed a relatively higher prediction performance in higher-order cognitive control than in primary sensorimotor systems. Furthermore, cognitive control systems exhibit a remarkable increase in prediction accuracy with tight task contrasts and a decrease in sensorimotor systems. This work demonstrates that probabilistic FC estimates are promising routes for activity flow mapping and also uncovers divergent influences of connectional topology and task contrasts on activity flow prediction across brain systems with different functional hierarchies.

A More Female-Characterized Resting-State Brain: Graph Similarity Analyses of Sex Influence on the Human Brain Intrinsic Functional Network.

It remains unclear whether human species exhibits sexual dimorphism in brain activities, and how the dimorphisms associated with sex-characterized behaviors. Here, in a large dataset from Human Connectome Project, we investigated sex differences of resting-state network structure by using local and global network graph similarity analysis. The "typical male" and "typical female" resting-state networks were highly similar. However, we found significant inter-sex difference in all local brain networks compared with sex-label permutations. The global and many local network topologies showed significant higher intra-female similarity, while males' network topologies were more dissimilar to each other. Additionally, by using global graph similarity analysis, we found that female individuals whose brain network were more similar to the average pattern present lower social-related anger, lower social distress and better companionships, while similar effects were not detected for males. Our study confirms the existence of sex-related resting-state network topology. Female's intrinsic brain is closer to a typical pattern than male's, and they may more fulfill the "similarity breeds connection" principle in building social ties.

Episodic memory in aspects of brain information transfer by resting-state network topology.

Cognitive functionality emerges due to neural interactions. The interregional signal interactions underlying episodic memory are a complex process. Thus, we need to quantify this process more accurately to understand how brain regions receive information from other regions. Studies suggest that resting-state functional connectivity (FC) conveys cognitive information; additionally, activity flow estimates the contribution of the source region to the activation pattern of the target region, thus decoding the cognitive information transfer. Therefore, we performed a combined analysis of task-evoked activation and resting-state FC voxel-wise by activity flow mapping to estimate the information transfer pattern of episodic memory. We found that the cinguloopercular (CON), frontoparietal (FPN) and default mode networks (DMNs) were the most recruited structures in information transfer. The patterns and functions of information transfer differed between encoding and retrieval. Furthermore, we found that information transfer was a better predictor of memory ability than previous methods. Additional analysis indicated that structural connectivity (SC) had a transportive role in information transfer. Finally, we present the information transfer mechanism of episodic memory from multiple neural perspectives. These findings suggest that information transfer is a better biological indicator that accurately describes signal communication in the brain and strongly influences the function of episodic memory.

Altered functional connectivity in newly diagnosed benign epilepsy with unilateral or bilateral centrotemporal spikes: A multi-frequency MEG study

ObjectiveRolandic epilepsy (RE) is one of the most common forms of epilepsy syndromes in children. The condition is usually accompanied with either unilateral or bilateral centrotemporal epileptic discharge. Despite the term “benign”, many studies have reported that children with benign epilepsy with centrotemporal spikes (BECTS) display a range of pervasive cognitive difficulties. In addition, existing research suggests that unilateral and bilateral centrotemporal spikes may affect cognition through different mechanisms. Consequently, the present study aimed to investigate cognitive impairment and the resting-state network topology of children with benign epilepsy with unilateral centrotemporal spikes (U-BECTS) and with bilateral centrotemporal spikes (B-BECTS). MethodsThis study recruited 14 children with U-BECTS and 14 with B-BECTS. Thereafter, cognition was assessed in 28 children with BECTS and 14 healthy controls, using the fourth edition of the Wechsler Intelligence Scale (WISC-IV). Additionally, the functional network of the brain was constructed through magnetoencephalography (MEG) to record the resting-state brain magnetic signals of the brain and by computing virtual sensor waveforms at the source level. Moreover, graph theory (GT) analysis was used to assess the properties of the brain network. ResultsChildren in the B-BECTS group had an earlier onset of epilepsy compared to those in the U-BECTS category. In addition, both the B-BECTS and U-BECTS groups had lower Full Scale Intelligence Quotient (FSIQ), Verbal Comprehension Index (VCI), and Working Memory Index (WMI) scores, compared to the healthy controls although only children in the B-BECTS category had lower Perceptual Reasoning Index (PRI) scores. The results also showed that both BECTS groups had increased frontal cortex connectivity in specific frequency bands. Notably, children with B-BECTS showed a more disorderly and randomized network in the 1–4-Hz and 80–250-Hz frequency bands. Moreover, GT analysis showed that children with B-BECTS had lower clustering coefficient and characteristic path length in the 80–250-Hz frequency bands and higher connection strength in the 4–8-Hz frequency bands. On the other hand, the U-BECTS group had a higher clustering coefficient in the 8–12-Hz frequency bands, compared to the healthy controls. Correlation analysis revealed that there were negative correlations between network parameters, clinical characteristics, and neuropsychological data in the U-BECTS category. ConclusionThe findings revealed that children with BECTS display a diffuse early cognitive deficit. In addition, resting-state suboptimal network topology may be the mechanism of cognitive impairment in children with BECTS. The study also showed that and children with B-BECTS may be at a higher risk of cognitive impairment.

Resting-state network topology characterizing callous-unemotional traits in adolescence

BackgroundCallous-unemotional (CU) traits, a youth antisocial phenotype, are hypothesized to associate with aberrant connectivity (dis-integration) across the salience (SAL), default mode (DMN), and frontoparietal (FPN) networks. However, CU traits have a heterogeneous presentation and previous research has not modeled individual heterogeneity in resting-state connectivity amongst adolescents with CU traits. The present study models individual-specific network maps and examines topological features of individual and subgroup maps in relation to CU traits. MethodsParticipants aged 13–17 (n = 84, male = 55%, female = 45%) completed resting-state functional connectivity and the inventory of callous-unemotional traits as part of the Nathan Klein Rockland study. A sparse network approach (GIMME) was used to derive individual-level and subgroup maps of all participants. We then examined heterogeneous network features, including positive and negative connection density, associated with CU traits. ResultsHigher rates of CU traits increased probability of inclusion in one subgroup, which had the highest mean level of CU traits. Analysis of network features reveals less density (positive and negative) within the FPN and greater density between DMN-FPN associated with CU traits. DiscussionFindings indicate heterogeneous person-specific connections and some subgroup connections amongst adolescents associate with CU traits. Higher CU traits associate with lower density in the FPN, which has been associated with attention and inhibition, and higher density between the DMN-FPN, which have been linked with cognitive control, social working memory, and empathy. Our findings suggest less efficiency in FPN function which, when considered mechanistically, could result in difficulty suppressing DMN when task positive networks are engaged. This is an area for further exploration but could explain cognitive and socio-affective impairments in CU traits.

Resting-state network topology and planning ability in healthy adults

Functional magnetic resonance imaging (fMRI) studies have been used extensively to investigate the brain areas that are recruited during the Tower of London (ToL) task. Nevertheless, little research has been devoted to study the neural correlates of the ToL task using a network approach. Here we investigated the association between functional connectivity and network topology during resting-state fMRI and ToL task performance, that was performed outside the scanner. Sixty-two (62) healthy subjects (21–74 years) underwent eyes-closed rsfMRI and performed the task on a laptop. We studied global (whole-brain) and within subnetwork resting-state topology as well as functional connectivity between subnetworks, with a focus on the default-mode, fronto-parietal and dorsal and ventral attention networks. Efficiency and clustering coefficient were calculated to measure network integration and segregation, respectively, at both the global and subnetwork level. Our main finding was that higher global efficiency was associated with slower performance (β = 0.22, Pbca = 0.04) and this association seemed mainly driven by inter-individual differences in default-mode network connectivity. The reported results were independent of age, sex, education-level and motion. Although this finding is contrary to earlier findings on general cognition, we tentatively hypothesize that the reported association may indicate that individuals with a more integrated brain during the resting-state are less able to further increase network efficiency when transitioning from a rest to task state, leading to slower responses. This study also adds to a growing body of literature supporting a central role for the default-mode network in individual differences in cognitive performance.

Altered resting-state network topology in drug-naïve idiopathic generalized epilepsy (P5.5-021)

Altered resting-state network topology in drug-naïve idiopathic generalized epilepsy (P5.5-021)

Clinical pain and functional network topology in Parkinson\u2019s disease: a resting-state fMRI study

Pain is an important non-motor symptom in Parkinson’s disease (PD), but its underlying pathophysiological mechanisms are still unclear. Research has shown that functional connectivity during the resting-state may be involved in persistent pain in PD. In the present cross-sectional study, 24 PD patients (both during on and off medication phase) and 27 controls participated. We assessed pain with the colored analogue scale and the McGill pain questionnaire. We examined a possible pathophysiological mechanism with resting-state fMRI using functional network topology, i.e., the architecture of functional connections. We took betweenness centrality (BC) to assess hubness, and global efficiency (GE) to assess integration of the network. We aimed to (1) assess the differences between PD patients and controls with respect to pain and resting-state network topology, and (2) investigate how resting-state network topology (BC and GE) is associated with clinical pain in both PD patients and controls. Results show that PD patients experienced more pain than controls. GE of the whole brain was higher in PD patients (on as well as off medication) compared to healthy controls. GE of the specialized pain network was also higher in PD patients compared to controls, but only when patients were on medication. BC of the pain network was lower in PD patients off medication compared to controls. We found a positive association between pain and GE of the pain network in PD patients off medication. For healthy controls, a negative association was found between pain and GE of the pain network, and also between pain and BC of the pain network. Our results suggest that functional network topology differs between PD patients and healthy controls, and that this topology can be used to investigate the underlying neural mechanisms of pain symptoms in PD.

Cognitive task information is transferred between brain regions via resting-state network topology

Resting-state network connectivity has been associated with a variety of cognitive abilities, yet it remains unclear how these connectivity properties might contribute to the neurocognitive computations underlying these abilities. We developed a new approach—information transfer mapping—to test the hypothesis that resting-state functional network topology describes the computational mappings between brain regions that carry cognitive task information. Here, we report that the transfer of diverse, task-rule information in distributed brain regions can be predicted based on estimated activity flow through resting-state network connections. Further, we find that these task-rule information transfers are coordinated by global hub regions within cognitive control networks. Activity flow over resting-state connections thus provides a large-scale network mechanism for cognitive task information transfer and global information coordination in the human brain, demonstrating the cognitive relevance of resting-state network topology.

Resting-State Network Topology Differentiates Task Signals across the Adult Life Span.

Brain network connectivity differs across individuals. For example, older adults exhibit less segregated resting-state subnetworks relative to younger adults (Chan et al., 2014). It has been hypothesized that individual differences in network connectivity impact the recruitment of brain areas during task execution. While recent studies have described the spatial overlap between resting-state functional correlation (RSFC) subnetworks and task-evoked activity, it is unclear whether individual variations in the connectivity pattern of a brain area (topology) relates to its activity during task execution. We report data from 238 cognitively normal participants (humans), sampled across the adult life span (20–89 years), to reveal that RSFC-based network organization systematically relates to the recruitment of brain areas across two functionally distinct tasks (visual and semantic). The functional activity of brain areas (network nodes) were characterized according to their patterns of RSFC: nodes with relatively greater connections to nodes in their own functional system (“non-connector” nodes) exhibited greater activity than nodes with relatively greater connections to nodes in other systems (“connector” nodes). This “activation selectivity” was specific to those brain systems that were central to each of the tasks. Increasing age was accompanied by less differentiated network topology and a corresponding reduction in activation selectivity (or differentiation) across relevant network nodes. The results provide evidence that connectional topology of brain areas quantified at rest relates to the functional activity of those areas during task. Based on these findings, we propose a novel network-based theory for previous reports of the “dedifferentiation” in brain activity observed in aging.SIGNIFICANCE STATEMENT Similar to other real-world networks, the organization of brain networks impacts their function. As brain network connectivity patterns differ across individuals, we hypothesized that individual differences in network connectivity would relate to differences in brain activity. Using functional MRI in a group of individuals sampled across the adult life span (20–89 years), we measured correlations at rest and related the functional connectivity patterns to measurements of functional activity during two independent tasks. Brain activity varied in relation to connectivity patterns revealed by large-scale network analysis. This relationship tracked the differences in connectivity patterns accompanied by older age, providing important evidence for a link between the topology of areal connectivity measured at rest and the functional recruitment of these areas during task performance.

Relationships between the resting-state network and the P3: Evidence from a scalp EEG study.

The P3 is an important event-related potential that can be used to identify neural activity related to the cognitive processes of the human brain. However, the relationships, especially the functional correlations, between resting-state brain activity and the P3 have not been well established. In this study, we investigated the relationships between P3 properties (i.e., amplitude and latency) and resting-state brain networks. The results indicated that P3 amplitude was significantly correlated with resting-state network topology, and in general, larger P3 amplitudes could be evoked when the resting-state brain network was more efficient. However, no significant relationships were found for the corresponding P3 latency. Additionally, the long-range connections between the prefrontal/frontal and parietal/occipital brain regions, which represent the synchronous activity of these areas, were functionally related to the P3 parameters, especially P3 amplitude. The findings of the current study may help us better understand inter-subject variation in the P3, which may be instructive for clinical diagnosis, cognitive neuroscience studies, and potential subject selection for brain-computer interface applications.

Prediction of SSVEP-based BCI performance by the resting-state EEG network

Objective. The prediction of brain–computer interface (BCI) performance is a significant topic in the BCI field. Some researches have demonstrated that resting-state data are promising candidates to achieve the goal. However, so far the relationships between the resting-state networks and the steady-state visual evoked potential (SSVEP)-based BCI have not been investigated. In this paper, we investigate the possible relationships between the SSVEP responses, the classification accuracy of five stimulus frequencies and the closed-eye resting-state network topology. Approach. The resting-state functional connectivity networks of the corresponding five stimulus frequencies were created by coherence, and then three network topology measures—the mean functional connectivity, the clustering coefficient and the characteristic path length of each network—were calculated. In addition, canonical correlation analysis was used to perform frequency recognition with the SSVEP data. Main results. Interestingly, we found that SSVEPs of each frequency were negatively correlated with the mean functional connectivity and clustering coefficient, but positively correlated with characteristic path length. Each of the averaged network topology measures across the frequencies showed the same relationship with the SSVEPs averaged across frequencies between the subjects. Furthermore, our results also demonstrated that the classification accuracy can be predicted by three averaged network measures and their combination can further improve the prediction performance. Significance. These findings indicate that the SSVEP responses and performance are predictable using the information at the resting-state, which may be instructive in both SSVEP-aided cognition studies and SSVEP-based BCI applications.

Interindividual reaction time variability is related to resting-state network topology: an electroencephalogram study

Both anatomical and functional brain network studies have drawn great attention recently. Previous studies have suggested the significant impacts of brain network topology on cognitive function. However, the relationship between non-task–related resting-state functional brain network topology and overall efficiency of sensorimotor processing has not been well identified. In the present study, we investigated the relationship between non-task–related resting-state functional brain network topology and reaction time (RT) in a Go/Nogo task using an electroencephalogram (EEG). After estimating the functional connectivity between each pair of electrodes, graph analysis was applied to characterize the network topology. Two fundamental measures, clustering coefficient (functional segregation) and characteristic path length (functional integration), as well as “small-world-ness” (the ratio between the clustering coefficient and characteristic path length) were calculated in five frequency bands. Then, the correlations between the network measures and RT were evaluated in each band separately. The present results showed that increased overall functional connectivity in alpha and gamma frequency bands was correlated with a longer RT. Furthermore, shorter RT was correlated with a shorter characteristic path length in the gamma band. This result suggested that human RTs were likely to be related to the efficiency of the brain integrating information across distributed brain regions. The results also showed that a longer RT was related to an increased gamma clustering coefficient and decreased small-world-ness. These results provided further evidence of the association between the resting-state functional brain network and cognitive function.