Traditional Functional Connectivity Research Articles

A benchmark of individual auto-regressive models in a massive fMRI dataset

Abstract Dense functional magnetic resonance imaging datasets open new avenues to create auto-regressive models of brain activity. Individual idiosyncrasies are obscured by group models, but can be captured by purely individual models given sufficient amounts of training data. In this study, we compared several deep and shallow individual models on the temporal auto-regression of BOLD time-series recorded during a natural video-watching task. The best performing models were then analyzed in terms of their data requirements and scaling, subject specificity, and the space-time structure of their predicted dynamics. We found the Chebnets, a type of graph convolutional neural network, to be best suited for temporal BOLD auto-regression, closely followed by linear models. Chebnets demonstrated an increase in performance with increasing amounts of data, with no complete saturation at 9 h of training data. Good generalization to other kinds of video stimuli and to resting-state data marked the Chebnets’ ability to capture intrinsic brain dynamics rather than only stimulus-specific autocorrelation patterns. Significant subject specificity was found at short prediction time lags. The Chebnets were found to capture lower frequencies at longer prediction time lags, and the spatial correlations in predicted dynamics were found to match traditional functional connectivity networks. Overall, these results demonstrate that large individual functional magnetic resonance imaging (fMRI) datasets can be used to efficiently train purely individual auto-regressive models of brain activity, and that massive amounts of individual data are required to do so. The excellent performance of the Chebnets likely reflects their ability to combine spatial and temporal interactions on large time scales at a low complexity cost. The non-linearities of the models did not appear as a key advantage. In fact, surprisingly, linear versions of the Chebnets appeared to outperform the original non-linear ones. Individual temporal auto-regressive models have the potential to improve the predictability of the BOLD signal. This study is based on a massive, publicly-available dataset, which can serve for future benchmarks of individual auto-regressive modeling.

Sliding window functional connectivity inference with nonstationary autocorrelations and cross-correlations.

Functional connectivity (FC) is the degree of synchrony of time series between distinct, spatially separated brain regions. While traditional FC analysis assumes the temporal stationarity throughout a brain scan, there is growing recognition that connectivity can change over time and is not stationary, leading to the concept of dynamic FC (dFC). Resting-state functional magnetic resonance imaging (fMRI) can assess dFC using the sliding window method with the correlation analysis of fMRI signals. Accurate statistical inference of sliding window correlation must consider the autocorrelated nature of the time series. Currently, the dynamic consideration is mainly confined to the point estimation of sliding window correlations. Using in vivo resting-state fMRI data, we first demonstrate the non-stationarity in both the cross-correlation function (XCF) and the autocorrelation function (ACF). Then, we propose the variance estimation of the sliding window correlation considering the nonstationary of XCF and ACF. This approach provides a means to dynamically estimate confidence intervals in assessing dynamic connectivity. Using simulations, we compare the performance of the proposed method with other methods, showing the impact of dynamic ACF and XCF on connectivity inference. Accurate variance estimation can help in addressing the critical issue of false positivity and negativity.

Electroencephalogram Functional Connectivity Analysis and Classification of Mental Arithmetic Working Memory Task

Analyzing brain activity during mental arithmetic tasks provides insight into psychological disorders such as ADHD, dyscalculia, and autism. While most research is conducted on the static functional connectivity of the brain while performing a cognitive task, the dynamic changes of the brain, which provide meaningful information for diagnosing individual differences in cognitive tasks, are often ignored. This paper aims to classify electroencephalogram (EEG) signals for rest vs. mental arithmetic task performance, using Bayesian functional connectivity features in the sensor space as inputs into a graph convolutional network. The subject-specific (intrasubject) classification performed on 36 subjects for rest vs. mental arithmetic task performance achieved the highest subject-specific classification accuracy of 98% and an average accuracy of 91% in the beta frequency band, outperforming state-of-the-art methods. In addition, statistical analysis confirms the consistency of Bayesian functional connectivity features compared to traditional functional connectivity features. Furthermore, the graph-theoretical analysis of functional connectivity networks reveals that good-performance subjects had higher global efficiency, betweenness centrality, and closeness centrality than bad-performance subjects. The ablation study on the classification of three cognitive states (subtraction, music, and memory) achieved a classification accuracy of 97%, and visual working memory (n-back task) achieved a classification accuracy of 94%, confirming the consistency and reliability of the proposed methodology.

Sustained connectivity correlates of motor function in Alzheimer’s disease

AbstractBackgroundIn addition to cognitive deficits, issues with motor control are present across the dementia spectrum. Traditional functional connectivity measures synchrony of intrinsic activity between brain regions and is altered in patients with Alzheimer’s disease in regions supporting motor function; however, less is known regarding the temporal stability of that synchrony. Sustained connectivity uses the width of cross correlation curves of BOLD time series to reflect the relative duration of synchronous activity between brain regions. This study investigates sustained connectivity across the dementia spectrum in motor regions.Method109 participants were included in this analysis and included 32 individuals with mild cognitive impairment (MCI) (18 F, 73.84 ± 5.14), 30 with Alzheimer’s disease (AD) (19 F, 77.37 ± 6.45), and 47 control individuals (29 F, mean age 72.13 ± 4.68). All study participants completed MP2RAGE structural imaging and two 15‐minute resting fMRI scans. Preprocessing steps included segmentation, motion realignment, registration to MNI space, and regression of motion parameters. Sustained connectivity values were averaged across both scans and only participants with more than 50% volumes remaining after censoring were included in analysis. A 17 resting‐state network parcellation was used to evaluate sustained connectivity across the brain with each network treated as a single ROI. A linear regression analysis was used to compare sustained connectivity across groups controlling for mean head motion, age, and sex. T statistics and p‐values are reported for across group sustained connectivity values.ResultAcross all three groups, sustained connectivity with the dorsal and ventral somatomotor networks was significantly increased across a number of resting‐state networks including the central visual, anterior ventral attention, and lateral temporal default mode network (DMN; all dorsal) and the somatomotor association, anterior ventral attention, medial frontoparietal, lateral temporal and lateral DMN (all ventral; Figure 1; pFDR < .05).ConclusionThese data suggest sustained connectivity may be a potential metric for assessing progressive cognitive decline in dementia. Our results indicate that sustained connectivity may be used to track changes in the temporal duration of functional connectivity in motor networks in individuals with intact cognition, mild cognitive impairment, and Alzheimer’s disease.

Fusing the spatial structure of electroencephalogram channels can increase the individualization of the functional connectivity network.

An electroencephalogram (EEG) functional connectivity (FC) network is individualized and plays a significant role in EEG-based person identification. Traditional FC networks are constructed by statistical dependence and correlation between EEG channels, without considering the spatial relationships between the channels. The individual identification algorithm based on traditional FC networks is sensitive to the integrity of channels and crucially relies on signal preprocessing; therefore, finding a new presentation for FC networks may help increase the performance of the identification algorithms. EEG signals are smooth across space owing to the volume conduction effect. Considering such spatial relationships among channels can provide a more accurate representation of FC networks. In this study, we propose an EEG FC network with virtual nodes that combines the spatial relationships and functional connectivity of channels. The comparison results for individual identification show that the novel EEG network is more individualized and achieves an accuracy of 98.64% for data without preprocessing. Furthermore, our algorithm is more robust in reducing the number of channels and can perform well even when a large area of channels is removed.

Test-retest reliability and predictive utility of a macroscale principal functional connectivity gradient.

Mapping individual differences in brain function has been hampered by poor reliability as well as limited interpretability. Leveraging patterns of brain-wide functional connectivity (FC) offers some promise in this endeavor. In particular, a macroscale principal FC gradient that recapitulates a hierarchical organization spanning molecular, cellular, and circuit level features along a sensory-to-association cortical axis has emerged as both a parsimonious and interpretable measure of individual differences in behavior. However, the measurement reliabilities of this FC gradient have not been fully evaluated. Here, we assess the reliabilities of both global and regional principal FC gradient measures using test-retest data from the young adult Human Connectome Project (HCP-YA) and the Dunedin Study. Analyses revealed that the reliabilities of principal FC gradient measures were (1) consistently higher than those for traditional edge-wise FC measures, (2) higher for FC measures derived from general FC (GFC) in comparison with resting-state FC, and (3) higher for longer scan lengths. We additionally examined the relative utility of these principal FC gradient measures in predicting cognition and aging in both datasets as well as the HCP-aging dataset. These analyses revealed that regional FC gradient measures and global gradient range were significantly associated with aging in all three datasets, and moderately associated with cognition in the HCP-YA and Dunedin Study datasets, reflecting contractions and expansions of the cortical hierarchy, respectively. Collectively, these results demonstrate that measures of the principal FC gradient, especially derived using GFC, effectively capture a reliable feature of the human brain subject to interpretable and biologically meaningful individual variation, offering some advantages over traditional edge-wise FC measures in the search for brain-behavior associations.

Abnormal higher-order network interactions in Parkinson's disease visual hallucinations.

Visual hallucinations in Parkinson's disease can be viewed from a systems-level perspective, whereby dysfunctional communication between brain networks responsible for perception predisposes a person to hallucinate. To this end, abnormal functional interactions between higher-order and primary sensory networks have been implicated in the pathophysiology of visual hallucinations in Parkinson's disease, however the precise signatures remain to be determined. Dimensionality reduction techniques offer a novel means for simplifying the interpretation of multidimensional brain imaging data, identifying hierarchical patterns in the data that are driven by both within- and between-functional network changes. Here, we applied two complementary non-linear dimensionality reduction techniques-diffusion-map embedding and t-distributed stochastic neighbour embedding (t-SNE)-to resting state functional MRI data, in order to characterize the altered functional hierarchy associated with susceptibility to visual hallucinations. Our study involved 77 people with Parkinson's disease (31 with hallucinations; 46 without hallucinations) and 19 age-matched healthy control subjects. In patients with visual hallucinations, we found compression of the unimodal-heteromodal gradient consistent with increased functional integration between sensory and higher order networks. This was mirrored in a traditional functional connectivity analysis, which showed increased connectivity between the visual and default mode networks in the hallucinating group. Together, these results suggest a route by which higher-order regions may have excessive influence over earlier sensory processes, as proposed by theoretical models of hallucinations across disorders. By contrast, the t-SNE analysis identified distinct alterations in prefrontal regions, suggesting an additional layer of complexity in the functional brain network abnormalities implicated in hallucinations, which was not apparent in traditional functional connectivity analyses. Together, the results confirm abnormal brain organization associated with the hallucinating phenotype in Parkinson's disease and highlight the utility of applying convergent dimensionality reduction techniques to investigate complex clinical symptoms. In addition, the patterns we describe in Parkinson's disease converge with those seen in other conditions, suggesting that reduced hierarchical differentiation across sensory-perceptual systems may be a common transdiagnostic vulnerability in neuropsychiatric disorders with perceptual disturbances.

Individual Proportion Loss of Functional Connectivity Strength: A Novel Individual Functional Connectivity Biomarker for Subjective Cognitive Decline Populations.

High individual variation in the subjective cognitive decline (SCD) population makes functional connectivity (FC) biomarkers unstable. This study proposed a novel individual FC index, named individual proportion loss of functional connectivity strength (IPLFCS), and explored potential biomarkers for SCD using this new index. We proposed an IPLFCS analysis framework and compared it with traditional FC in Chinese and Western cohorts. Post hoc tests were used to determine biomarkers. Pearson's correlation analysis was used to investigate the correlation between neuropsychological scores or cortical amyloid deposits and IPLFCS biomarkers. Receiver operating characteristic curves were utilized to evaluate the ability of potential biomarkers to distinguish between groups. IPLFCS of the left middle temporal gyrus (LMTG) was identified as a potential biomarker. The IPLFC was correlated with the traditional FC (r = 0.956, p < 0.001; r = 0.946, p < 0.001) and cortical amyloid deposition (r = -0.245, p = 0.029; r = -0.185, p = 0.048) in both cohorts. Furthermore, the IPLFCS decreased across the Alzheimer's disease (AD) continuum. Its diagnostic efficiency was superior to that of existing fMRI biomarkers. These findings suggest that IPLFCS of the LMTG could be a potential biomarker of SCD.

Recurrence quantification analysis of periodic dynamics in the default mode network in first-episode drug-naïve schizophrenia

Abnormal functional connectivity (FC) within the default model network (DMN) in schizophrenia has been frequently reported in previous studies. However, traditional FC analysis was mostly linear correlations based, with the information on nonlinear or temporally lagged brain signals largely overlooked. Fifty-five first-episode drug-naïve schizophrenia (FES) patients and 53 healthy controls (HCs) underwent resting-state functional magnetic resonance imaging scanning. The DMN was extracted using independent component analysis. Recurrence quantification analysis was used to measure the duration, predictability, and complexity of the periodic processes of the nonlinear DMN time series. The Mann‒Whitney U test was conducted to compare these features between FES patients and HCs. The support vector machine was applied to discriminate FES from HCs based on these features. Determinism, which means predictability of periodic process activity, between the ventromedial prefrontal cortex (vMPFC) and posterior cingulate and between the vMPFC and precuneus, was significantly decreased in FES compared with HCs. Determinism between the vMPFC and precuneus was positively correlated with category fluency scores in FES. The classifier achieved 77% accuracy. Our results suggest that synchronized periodicity among DMN brain regions is dysregulated in FES, and the periodicity in BOLD signals may be a promising indicator of brain functional connectivity.

Effective Connectivity Is A More Promising Biomarker for Brain Health in Middle‐Aged Adults Than Functional Connectivity: Bogalusa Heart Study

AbstractBackgroundEffective connectivity, the causal influence that functional activity in a source brain location exerts over functional activity in a target brain location, has the potential to provide richer information about brain networks than functional connectivity, which only quantifies activity synchrony between regions. However, head‐to‐head comparisons of effective and functional connectivity from functional MRI data are rare, especially among diverse cognitively healthy middle aged and older adults.Method100 Bogalusa Heart Study participants performed a Stroop task during functional MRI (Table 1). Effective connectivity and functional connectivity among 24 regions of interest (ROIs) previously identified as activated by this task [1] were calculated using deep stacking networks [2] and Pearson correlation, respectively. Graph metrics were then calculated from directed graphs resulting from effective connectivity and undirected graphs resulting from functional connectivity. Linear regression models related graph metrics to demographic, cardiometabolic, and cognitive measures.ResultFigure 1 shows the group‐consensus effective and functional connectivity graphs. Systolic blood pressure (SBP) was associated with effective connectivity based characteristic path length (Pearson’s r = 0.22, p‐value = 0.038) and modularity (r = 0.25, p‐value = 0.02) (Figure 2). Diastolic blood pressure (DBP) was associated with effective connectivity based degree, density, clustering coefficient, modularity, transitivity, characteristic path length, and global efficiency (|r| = 0.26 ‐ 0.35, maximum p‐value = 0.013) (Figure 2). Word reading and vocabulary scores from the Wide Range Achievement Test (WRAT) were associated with effective connectivity based degree, density, clustering coefficient, transitivity, characteristic path length, modularity, and global efficiency (|r| = 0.21 ‐ 0.27, maximum p‐value = 0.049). Logical Memory II Recognition scores associated with effective connectivity based assortativity (r = 0.24, p‐value = 0.025). Associations between functional connectivity based graph metrics and demographic, cardiometabolic, and cognitive measures were not significant.ConclusionIn a diverse, cognitively healthy, middle‐aged community sample, graph metrics derived from effective connectivity based directed graphs tracked more closely with recognized indicators of brain health than traditional functional connectivity based undirected graph metrics.

Effective Connectivity Is A More Promising Biomarker for Brain Health in Middle‐Aged Adults Than Functional Connectivity: Bogalusa Heart Study

AbstractBackgroundEffective connectivity, the causal influence that functional activity in a source brain location exerts over functional activity in a target brain location, has the potential to provide richer information about brain networks than functional connectivity, which only quantifies activity synchrony between regions. However, head‐to‐head comparisons of effective and functional connectivity from functional MRI data are rare, especially among diverse cognitively healthy middle aged and older adults.Method100 Bogalusa Heart Study participants performed a Stroop task during functional MRI (Table 1). Effective connectivity and functional connectivity among 24 regions of interest (ROIs) previously identified as activated by this task [1] were calculated using deep stacking networks [2] and Pearson correlation, respectively. Graph metrics were then calculated from directed graphs resulting from effective connectivity and undirected graphs resulting from functional connectivity. Linear regression models related graph metrics to demographic, cardiometabolic, and cognitive measures.ResultFigure 1 shows the group‐consensus effective and functional connectivity graphs. Systolic blood pressure (SBP) was associated with effective connectivity based characteristic path length (Pearson’s r = 0.22, p‐value = 0.038) and modularity (r = 0.25, p‐value = 0.02) (Figure 2). Diastolic blood pressure (DBP) was associated with effective connectivity based degree, density, clustering coefficient, modularity, transitivity, characteristic path length, and global efficiency (|r| = 0.26 ‐ 0.35, maximum p‐value = 0.013) (Figure 2). Word reading and vocabulary scores from the Wide Range Achievement Test (WRAT) were associated with effective connectivity based degree, density, clustering coefficient, transitivity, characteristic path length, modularity, and global efficiency (|r| = 0.21 ‐ 0.27, maximum p‐value = 0.049). Logical Memory II Recognition scores associated with effective connectivity based assortativity (r = 0.24, p‐value = 0.025). Associations between functional connectivity based graph metrics and demographic, cardiometabolic, and cognitive measures were not significant.ConclusionIn a diverse, cognitively healthy, middle‐aged community sample, graph metrics derived from effective connectivity based directed graphs tracked more closely with recognized indicators of brain health than traditional functional connectivity based undirected graph metrics.

Blood Oxygen Level-Dependent Lag Patterns Differ Between Rest and Task Conditions, but Are Largely Typical in Autism.

Background/Introduction: Autism spectrum disorder (ASD) is characterized by atypical functional connectivity (FC) within and between distributed brain networks. However, FC findings have often been inconsistent, possibly due to a focus on static FC rather than brain dynamics. Lagged connectivity analyses aim at evaluating temporal latency, and presumably neural propagation, between regions. This approach may, therefore, reveal a more detailed picture of network organization in ASD than traditional FC methods. Methods: The current study evaluated whole-brain lag patterns in adolescents with ASD (n = 28) and their typically developing peers (n = 22). Functional magnetic resonance imaging data were collected during rest and during a lexico-semantic decision task. Optimal lag was calculated for each pair of regions of interest by using cross-covariance, and mean latency projections were calculated for each region. Results: Latency projections did not regionally differ between groups, with the same regions emerging among the "earliest" and "latest." Although many of the longest absolute latencies were preserved across resting-state and task conditions, lag patterns overall were affected by condition, as many regions shifted toward zero-lag during task performance. Lag structure was also strongly associated with literature-derived estimates of arterial transit time. Discussion: Results suggest that lag patterns are broadly typical in ASD but undergo changes during task performance. Moreover, lag patterns appear to reflect a combination of neural and vascular sources, which should be carefully considered when interpreting lagged FC. Impact statement Altered brain dynamics have been proposed in autism spectrum disorder (ASD). Lagged functional connectivity analysis uses cross-correlation between functional magnetic resonance imaging (fMRI) time series to determine regional latency. Few studies have examined blood oxygen level-dependent (BOLD) lag in ASD, and findings have been inconsistent. Using multi-echo fMRI data with improved artifact detection and removal, we find differences in lag structure between task and rest states, but not between adolescents with ASD and typically developing peers. Additional analyses exploring links with arterial transit time, however, highlight the impact of vascular organization on BOLD lag patterns and its potential to confound measures of neural dynamics.

Dynamic Time Warping Identifies Functionally Distinct fMRI Resting State Cortical Networks Specific to VTA and SNc: A Proof of Concept.

Functional connectivity (FC) is determined by similarity between functional magnetic resonance imaging (fMRI) signals from distinct brain regions. However, traditional FC analyses ignore temporal phase differences. Here, we addressed this limitation, using dynamic time warping (DTW) within a machine-learning framework, to study cortical FC patterns of 2 spatially adjacent but functionally distinct subcortical regions, namely Substantia Nigra Pars Compacta (SNc) and ventral tegmental area (VTA). We evaluate: 1) the influence of pair of brain regions considered, 2) the influence of warping window sizes, 3) the classification efficacy of DTW, and 4) the uniqueness of features identified. Whole brain 7 Tesla resting state fMRI scans from 81 healthy participants were used. FC between 2 subcortical regions of interests (ROIs) and 360 cortical parcels were computed using: 1) Pearson correlations (PCs), 2) dynamic time-warped PCs (DTW-PC). The separability of SNc-cortical and VTA-cortical network was validated on 40 participants and tested on the remaining 41, using a support vector machine (SVM). The SVM separated the SNc-cortical versus VTA-cortical network with 74.39 and 97.56% test accuracy using PC and DTW-PC, respectively. SVM-recursive feature elimination yielded 20 DTW-PC features that most strongly contributed to the separation of the networks and revealed novel VTA versus SNc preferential connections (P < 0.05, Bonferroni-Holm corrected).

Obese Individuals Show Disrupted Dynamic Functional Connectivity between Basal Ganglia and Salience Networks.

Previous functional magnetic resonance imaging (fMRI) studies have showed obesity (OB)-related alterations in intrinsic functional connectivity (FC) within and between different resting-state networks (RSNs). However, few studies have examined dynamic functional connectivity (DFC). Thus, we employed resting-state fMRI with independent component analysis (ICA) and DFC analysis to investigate the alterations in FC within and between RSNs in 56 individuals with OB and 46 normal-weight (NW) controls. ICA identified six RSNs, including basal ganglia (BG), salience network (SN), right executive control network/left executive control network, and anterior default-mode network (aDMN)/posterior default-mode network. The DFC analysis identified four FC states. OB compared with NW had more occurrences and a longer mean dwell time (MDT) in state 2 (positive connectivity of BG with other RSN) and also had higher FC of BG-SN in other states. Body mass index was positively correlated with MDT and FCs of BG-aDMN (state 2) and BG-SN (state 4). DFC analysis within more refined nodes of RSNs showed that OB had more occurrences and a longer MDT in state 1 in which caudate had positive connections with the other network nodes. The findings suggest an association between caudate-related and BG-related positive FC in OB, which was not revealed by traditional FC analysis, highlighting the utility of adding DFC to the more conventional methods.

T134. THE ROLE OF THE DEFAULT MODE NETWORK IN SCHIZOPHRENIA AND AUDITORY VERBAL HALLUCINATIONS – AN INVESTIGATION OF DYNAMIC FMRI RESTING STATE CONNECTIVITY

BackgroundThere is a wealth of evidence showing aberrant functional connectivity (FC) in schizophrenia but with considerable variability in findings across studies. Dynamic FC is an extension of traditional static FC, in that such analyses allow for explorations of temporal changes in connectivity. Thereby they also provide more detailed information on connectivity abnormalities in psychiatric disorders such as schizophrenia.MethodsThe current study investigated dynamic FC in a sample of 80 schizophrenia patients and 80 matched healthy control subjects. Furthermore, relationships with auditory verbal hallucinations (AVH), a core symptom of schizophrenia, were explored. Two measures of AVH were used, one measure of current AVH severity assessed on the day of scanning, and one trait-measure where AVH were assessed repeatedly over the course of one year.ResultsCompared to healthy controls, schizophrenia patients showed increased dwell times in states with high connectivity within the default mode network (DMN). Current AVH severity did not show a significant relationship with dynamic FC. However, the trait-measure of AVH proneness over one year showed a significant relationship with dynamic FC. Patients with high AVH proneness spent less time in connectivity states characterized by strong anti-correlation between the DMN and task-positive networks.DiscussionThe results provide further evidence for a DMN dysfunction in schizophrenia, which could be linked to thought disturbances in relation to an increased internal focus of cognitive processing. The effects of AVH proneness on dynamic FC support theoretical models of AVH which have proposed an instability of the DMN and impaired cognitive control in AVH patients. The results also point to AVH proneness as a potential marker for identifying distinct subgroups of schizophrenia patients.

Multi-Class ASD Classification Based on Functional Connectivity and Functional Correlation Tensor via Multi-Source Domain Adaptation and Multi-View Sparse Representation.

The resting-state functional magnetic resonance imaging (rs-fMRI) reflects functional activity of brain regions by blood-oxygen-level dependent (BOLD) signals. Up to now, many computer-aided diagnosis methods based on rs-fMRI have been developed for Autism Spectrum Disorder (ASD). These methods are mostly the binary classification approaches to determine whether a subject is an ASD patient or not. However, the disease often consists of several sub-categories, which are complex and thus still confusing to many automatic classification methods. Besides, existing methods usually focus on the functional connectivity (FC) features in grey matter regions, which only account for a small portion of the rs-fMRI data. Recently, the possibility to reveal the connectivity information in the white matter regions of rs-fMRI has drawn high attention. To this end, we propose to use the patch-based functional correlation tensor (PBFCT) features extracted from rs-fMRI in white matter, in addition to the traditional FC features from gray matter, to develop a novel multi-class ASD diagnosis method in this work. Our method has two stages. Specifically, in the first stage of multi-source domain adaptation (MSDA), the source subjects belonging to multiple clinical centers (thus called as source domains) are all transformed into the same target feature space. Thus each subject in the target domain can be linearly reconstructed by the transformed subjects. In the second stage of multi-view sparse representation (MVSR), a multi-view classifier for multi-class ASD diagnosis is developed by jointly using both views of the FC and PBFCT features. The experimental results using the ABIDE dataset verify the effectiveness of our method, which is capable of accurately classifying each subject into a respective ASD sub-category.

Dynamic Functional Connectivity Patterns in Schizophrenia and the Relationship With Hallucinations.

There is a wealth of evidence showing aberrant functional connectivity (FC) in schizophrenia but with considerable variability in findings across studies. Dynamic FC is an extension of traditional static FC, in that such analyses allow for explorations of temporal changes in connectivity. Thereby they also provide more detailed information on connectivity abnormalities in psychiatric disorders such as schizophrenia. The current study investigated dynamic FC in a sample of 80 schizophrenia patients and 80 matched healthy control subjects, replicating previous findings of aberrant dwell times in specific FC states, and further supporting a role for default mode network (DMN) dysfunction. Furthermore, relationships with hallucinations, a core symptom of schizophrenia, were explored. Two measures of hallucinations were used, one measure of current hallucination severity assessed on the day of scanning, and one trait-measure where hallucinations were assessed repeatedly over the course of 1 year. Current hallucination severity did not show a significant relationship with dynamic FC. However, the trait-measure of hallucination proneness over 1 year showed a significant relationship with dynamic FC. Patients with high hallucination proneness spent less time in connectivity states characterized by strong anti-correlation between the DMN and task-positive networks. The findings support theoretical models of hallucinations which have proposed an instability of the DMN and impaired cognitive control in patients with hallucinations. Furthermore, the results point to hallucination proneness as a potential marker for identifying distinct subgroups of schizophrenia patients.

Omission of temporal nuisance regressors from dual regression can improve accuracy of fMRI functional connectivity maps.

Functional connectivity (FC) maps from brain fMRI data can be derived with dual regression, a proposed alternative to traditional seed-based FC (SFC) methods that detect temporal correlation between a predefined region (seed) and other regions in the brain. As with SFC, incorporating nuisance regressors (NR) into the dual regression must be done carefully, to prevent potential bias and insensitivity of FC estimates. Here, we explore the potentially untoward effects on dual regression that may occur when NR correlate highly with the signal of interest, using both synthetic and real fMRI data to elucidate mechanisms responsible for loss of accuracy in FC maps. Our tests suggest significantly improved accuracy in FC maps derived with dual regression when highly correlated temporal NR were omitted. Single-map dual regression, a simplified form of dual regression that uses neither spatial nor temporal NR, offers a viable alternative whose FC maps may be more easily interpreted, and in some cases be more accurate than those derived with standard dual regression.

Fusion of ULS Group Constrained High- and Low-Order Sparse Functional Connectivity Networks for MCI Classification.

Functional connectivity networks, derived from resting-state fMRI data, have been found as effective biomarkers for identifying mild cognitive impairment (MCI) from healthy elderly. However, the traditional functional connectivity network is essentially a low-order network with the assumption that the brain activity is static over the entire scanning period, ignoring temporal variations among the correlations derived from brain region pairs. To overcome this limitation, we proposed a new type of sparse functional connectivity network to precisely describe the relationship of temporal correlations among brain regions. Specifically, instead of using the simple pairwise Pearson's correlation coefficient as connectivity, we first estimate the temporal low-order functional connectivity for each region pair based on an ULS Group constrained-UOLS regression algorithm, where a combination of ultra-least squares (ULS) criterion with a Group constrained topology structure detection algorithm is applied to detect the topology of functional connectivity networks, aided by an Ultra-Orthogonal Least Squares (UOLS) algorithm to estimate connectivity strength. Compared to the classical least squares criterion which only measures the discrepancy between the observed signals and the model prediction function, the ULS criterion takes into consideration the discrepancy between the weak derivatives of the observed signals and the model prediction function and thus avoids the overfitting problem. By using a similar approach, we then estimate the high-order functional connectivity from the low-order connectivity to characterize signal flows among the brain regions. We finally fuse the low-order and the high-order networks using two decision trees for MCI classification. Experimental results demonstrate the effectiveness of the proposed method on MCI classification.

Temporal-Spatial Patterns in Dynamic Functional Brain Network for Self-Paced Hand Movement.

Dynamic functional connectivity is attracting a growing interest as it has been suggested to be a more accurate representation of functional brain networks compared to traditional functional connectivity. It is believed that the functional connectivity fluctuations result from the transitions among different brain states other than continuous changes in the brain. In this paper, we aim to investigate the spatial-temporal changes in the interactions between different brain regions during a self-paced hand movement with EEG signals. A systematic analysis framework, consisting of connectivity metric calculation, brain state segmentation, temporal representative graph extraction, and spatial community detection, is proposed to analyze the dynamic functional connectivity. First, corrected imaginary coherency is applied to measure the functional connectivity as it is insensitive to EEG volume conduction problem. Second, singular value decomposition (SVD) vector space distance between the connectivity matrices at two adjacent time points is calculated. In addition, the brain states are segmented based on the changes in the time series of SVD vector space distances. Third, one representative graph is summarized within each state segment using the SVD vectors corresponding to the k largest singular values. Finally, spatial patterns on the representative graph are detected with a modularity-based community detection method. Based on the SVD vector space distance using the change point detection method, a series of brain states lasting for hundreds of milliseconds are identified. Moreover, we find that the sudden decrease points in SVD vector space distance coincide with early Bereitschafts potential. In addition, we find that there are several connectivity patterns along the time before the onset of movement. At first, the functional connectivity is relatively dispersed. Gradually, the functional connectivity begins to concentrate and the predominant communities in each dynamic functional network can be observed clearly.