Fluctuations In Brain Activity Research Articles

Distinctive time-lagged resting-state networks revealed by simultaneous EEG-fMRI

Functional activation as evidenced by blood oxygen level-dependent (BOLD) functional MRI changes or event-related EEG is known to closely follow patterns of stimulation or self-paced action. Any lags are compatible with axonal conduction velocities and neural integration times. The important analysis of resting state networks is generally based on the assumption that these principles also hold for spontaneous fluctuations in brain activity. Previous observations using simultaneous EEG and fMRI indicate that slower processes, with delays in the seconds range, determine at least part of the relationship between spontaneous EEG and fMRI. To assess this relationship systematically, we used deconvolution analysis of EEG-fMRI during the resting state, assessing the relationship between EEG frequency bands and fMRI BOLD across the whole brain while allowing for time lags of up to 10.5s. Cluster analysis, identifying similar BOLD time courses in relation to EEG band power peaks, showed a clear segregation of functional subsystems of the brain. Our analysis shows that fMRI BOLD increases commonly precede EEG power increases by seconds. Most zero-lag correlations, on the other hand, were negative. This indicates two main distinct neuromodulatory mechanisms: an “idling” mechanism of simultaneous electric and metabolic network anticorrelation and a “regulatory” mechanism in which metabolic network activity precedes increased EEG power by some seconds. This has to be taken into consideration in further studies which address the causal and functional relationship of metabolic and electric brain activity patterns.

Fluctuations in Global Brain Activity Are Associated With Changes in Whole-Brain Connectivity of Functional Networks

The aim of this study was to explore the relationship between global brain activity, changes in whole-brain connectivity, and changes in brain states across subjects using resting-state functional magnetic resonance imaging. We extended current methods that use a sparse set of coactivation patterns to extract critical time points in global brain activity. Critical activity time points were defined as points where the global signal is greater than one standard deviation above or below the average global signal. Four categories of critical points were defined along dimensions of global signal intensity and trajectory. Voxel-based methods were used to interrogate differences in connectivity between these critical points. Several differences in connectivity were found in functional resting-state networks (RSNs) as a function of global activity. RSNs associated with cognitive functions in frontal, parietal, and subcortical regions exhibited greater whole-brain connectivity during lower global activity states. Meanwhile, RSNs associated with sensory functions exhibited greater whole-brain connectivity during the higher global activity states. Moreover, we present evidence that these results depend in part upon the standard deviation threshold used to define the critical points, suggesting critical points at different thresholds represent unique brain states. Overall, the findings support the hypothesis that the brain oscillates through different states over the course of a resting-state study reflecting differences in RSN connectivity associated with global brain activity. Increased understanding of brain dynamics may help to elucidate individual differences in behavior and dysfunction.

The suppression of scale-free fMRI brain dynamics across three different sources of effort: aging, task novelty and task difficulty.

There is growing evidence that fluctuations in brain activity may exhibit scale-free (“fractal”) dynamics. Scale-free signals follow a spectral-power curve of the form P(f ) ∝ f−β, where spectral power decreases in a power-law fashion with increasing frequency. In this study, we demonstrated that fractal scaling of BOLD fMRI signal is consistently suppressed for different sources of cognitive effort. Decreases in the Hurst exponent (H), which quantifies scale-free signal, was related to three different sources of cognitive effort/task engagement: 1) task difficulty, 2) task novelty, and 3) aging effects. These results were consistently observed across multiple datasets and task paradigms. We also demonstrated that estimates of H are robust across a range of time-window sizes. H was also compared to alternative metrics of BOLD variability (SDBOLD) and global connectivity (Gconn), with effort-related decreases in H producing similar decreases in SDBOLD and Gconn. These results indicate a potential global brain phenomenon that unites research from different fields and indicates that fractal scaling may be a highly sensitive metric for indexing cognitive effort/task engagement.

Assessing the mean strength and variations of the time-to-time fluctuations of resting-state brain activity.

The time-to-time fluctuations (TTFs) of resting-state brain activity as captured by resting-state fMRI (rsfMRI) have been repeatedly shown to be informative of functional brain structures and disease-related alterations. TTFs can be characterized by the mean and the range of successive difference. The former can be measured with the mean squared successive difference (MSSD), which is mathematically similar to standard deviation; the latter can be calculated by the variability of the successive difference (VSD). The purpose of this study was to evaluate both the resting state-MSSD and VSD of rsfMRI regarding their test-retest stability, sensitivity to brain state change, as well as their biological meanings. We hypothesized that MSSD and VSD are reliable in resting brain; both measures are sensitive to brain state changes such as eyes-open compared to eyes-closed condition; both are predictive of age. These hypotheses were tested with three rsfMRI datasets and proven true, suggesting both MSSD and VSD as reliable and useful tools for resting-state studies.

Dynamic Multiscale Modes of Resting State Brain Activity Detected by Entropy Field Decomposition.

The ability of functional magnetic resonance imaging (FMRI) to noninvasively measure fluctuations in brain activity in the absence of an applied stimulus offers the possibility of discerning functional networks in the resting state of the brain. However, the reconstruction of brain networks from these signal fluctuations poses a significant challenge because they are generally nonlinear and nongaussian and can overlap in both their spatial and temporal extent. Moreover, because there is no explicit input stimulus, there is no signal model with which to compare the brain responses. A variety of techniques have been devised to address this problem, but the predominant approaches are based on the presupposition of statistical properties of complex brain signal parameters, which are unprovable but facilitate the analysis. In this article, we address this problem with a new method, entropy field decomposition, for estimating structure within spatiotemporal data. This method is based on a general information field-theoretic formulation of Bayesian probability theory incorporating prior coupling information that allows the enumeration of the most probable parameter configurations without the need for unjustified statistical assumptions. This approach facilitates the construction of brain activation modes directly from the spatial-temporal correlation structure of the data. These modes and their associated spatial-temporal correlation structure can then be used to generate space-time activity probability trajectories, called functional connectivity pathways, which provide a characterization of functional brain networks.

State dependency of inhibitory control performance: an electrical neuroimaging study.

Behavioral and brain responses to stimuli not only depend on their physical features but also on the individuals' neurocognitive states before stimuli onsets. While the influence of pre-stimulus fluctuations in brain activity on low-level perceptive processes is well established, the state dependency of high-order executive processes remains unclear. Using a classical inhibitory control Go/NoGo task, we examined whether and how fluctuations in the brain activity during the period preceding the stimuli triggering inhibition influenced inhibitory control performance. Seventeen participants completed the Go/NoGo task while 64-channel electroencephalogram was recorded. We compared the event-related potentials preceding the onset of the NoGo stimuli associated with inhibition failures false alarms (FA) vs. successful inhibition correct rejections (CR) with data-driven statistical analyses of global measures of the topography and strength of the scalp electric field. Distributed electrical source estimations were used to localize the origin of the event-related potentials modulations. We observed differences in the global field power of the event-related potentials (FA>CR) without concomitant topographic modulations over the 40ms period immediately preceding NoGo stimuli. This result indicates that the same brain networks were engaged in the two conditions, but more strongly before FA than CR. Source estimations revealed that this effect followed from a higher activity before FA than CR within bilateral inferior frontal gyri and the right inferior parietal lobule. These findings suggest that uncontrolled quantitative variations in pre-stimulus activity within attentional and control brain networks influence inhibition performance. The present data thereby demonstrate the state dependency of cognitive processes of up to high-order executive levels.

Reactivation of visual-evoked activity in human cortical networks.

In the absence of sensory input, neuronal networks are far from being silent. Whether spontaneous changes in ongoing activity reflect previous sensory experience or stochastic fluctuations in brain activity is not well understood. Here we demonstrate reactivation of stimulus-evoked activity that is distributed across large areas in the human brain. We performed simultaneous electrocorticography recordings from occipital, parietal, temporal, and frontal areas in awake humans in the presence and absence of sensory stimulation. We found that, in the absence of visual input, repeated exposure to brief natural movies induces robust stimulus-specific reactivation at individual recording sites. The reactivation sites were characterized by greater global connectivity compared with those sites that did not exhibit reactivation. Our results indicate a surprising degree of short-term plasticity across multiple networks in the human brain as a result of repeated exposure to unattended information.

Altered resting-state whole-brain functional networks of neonates with intrauterine growth restriction

The feasibility to use functional MRI (fMRI) during natural sleep to assess low-frequency basal brain activity fluctuations in human neonates has been demonstrated, although its potential to characterise pathologies of prenatal origin has not yet been exploited. In the present study, we used intrauterine growth restriction (IUGR) as a model of altered neurodevelopment due to prenatal condition to show the suitability of brain networks to characterise functional brain organisation at neonatal age. Particularly, we analysed resting-state fMRI signal of 20 neonates with IUGR and 13 controls, obtaining whole-brain functional networks based on correlations of blood oxygen level-dependent (BOLD) signal in 90 grey matter regions of an anatomical atlas (AAL). Characterisation of the networks obtained with graph theoretical features showed increased network infrastructure and raw efficiencies but reduced efficiency after normalisation, demonstrating hyper-connected but sub-optimally organised IUGR functional brain networks. Significant association of network features with neurobehavioral scores was also found. Further assessment of spatiotemporal dynamics displayed alterations into features associated to frontal, cingulate and lingual cortices. These findings show the capacity of functional brain networks to characterise brain reorganisation from an early age, and their potential to develop biomarkers of altered neurodevelopment.

Cognitive training can reduce the rate of cognitive aging: a neuroimaging cohort study.

BackgroundThe neural mechanisms underlying the restorative effects of cognitive training on aging brains remain unclear. To address this issue, we examined the relationship between changes in spontaneous brain activity and cognitive performance that occur after cognitive training.MethodsParticipants were older adults who were part of a randomized control trial within a larger longitudinal cognitive training study. We conducted single-domain and multi-domain cognitive training in two respective intervention groups. Participants were trained for 1 h, twice a week, for 12 weeks. Cognition was assessed in all participants and magnetic resonance images were obtained at baseline and 1 year after training. To assess spontaneous fluctuations in brain activity, we acquired resting-state fMRI data. Two indices—functional entropy and time-domain entropy—were used to measure the effects of training. Functional entropy increases with aging, and indicates disruptions in functional conectivity. Time-domain entropy decreases with aging, and indicates structural alterations in the brain and blood-flow reduction.ResultsSeventy participants completed the study: 26 in the multi-domain cognitive training group (70.38 ± 3.30 yrs), 27 in single-domain group (70.48 ± 3.93 yrs), and 17 in a control group (68.59 ± 3.24 yrs). Functional entropy increased significantly less in the multi-domain (p = 0.047) and single-domain groups (p = 9.51 × 10−4) compared with the control group. In the multi-domain group, this was true in the paracentral lobule (p = 0.004, Bonferroni corrected p < 0.05). Time-domain entropy also improved with training. Compared with controls, time-domain entropy in the multi-domain group decreased less in the inferior frontal gyrus pars opercularis (p = 3.59 × 10−4), the medial part of superior frontal gyrus (p = 1.17 × 10−5), and the thalamus (p = 4.72 × 10−5), while that in the single-domain group decreased less in the cuneus (p = 2.58 × 10−4, Bonferroni corrected p < 0.05). Additionally, changes in regional entropy for some regions such as hippocampus significantly correlated with improvements in cognitive performance.ConclusionsCognitive training can induce plastic changes in neural functional connectivity of healthy older people, and these changes may underlie the positive effect of cognitive training.Trial registrationChiCTR-TRC-08000732 (Date of registration: 5th November, 2008).Electronic supplementary materialThe online version of this article (doi:10.1186/s12877-016-0194-5) contains supplementary material, which is available to authorized users.

Spontaneous Fluctuations in the Flexible Control of Covert Attention.

Individuals regularly experience fluctuations in preparatory cognitive control that affect performance in everyday life. For example, individuals are able to more quickly initiate a spatial shift of attention at some moments than at others. The current study revealed that pretrial brain activity in specific cortical regions predicted trial-by-trial changes in participants' abilities to flexibly shift the focus of attention. Intrinsically generated fluctuations in brain activity within several key default mode network regions, as well as within the anterior insula and presupplementary/supplementary motor areas, carried behavioral consequences for preparatory attentional control beyond lapses of attentional engagement. Our results are the first to link intrinsic variation in pretrial brain activity to moment-by-moment changes in preparatory attentional control over spatial selection.

Neuronal Substrates Underlying Performance Variability in Well-Trained Skillful Motor Task in Humans

Motor performance fluctuates trial by trial even in a well-trained motor skill. Here we show neural substrates underlying such behavioral fluctuation in humans. We first scanned brain activity with functional magnetic resonance imaging while healthy participants repeatedly performed a 10 s skillful sequential finger-tapping task. Before starting the experiment, the participants had completed intensive training. We evaluated task performance per trial (number of correct sequences in 10 s) and depicted brain regions where the activity changes in association with the fluctuation of the task performance across trials. We found that the activity in a broader range of frontoparietocerebellar network, including the bilateral dorsolateral prefrontal cortex (DLPFC), anterior cingulate and anterior insular cortices, and left cerebellar hemisphere, was negatively correlated with the task performance. We further showed in another transcranial direct current stimulation (tDCS) experiment that task performance deteriorated, when we applied anodal tDCS to the right DLPFC. These results indicate that fluctuation of brain activity in the nonmotor frontoparietocerebellar network may underlie trial-by-trial performance variability even in a well-trained motor skill, and its neuromodulation with tDCS may affect the task performance.

Resting-state low-frequency fluctuations reflect individual differences in spoken language learning

A major challenge in language learning studies is to identify objective, pre-training predictors of success. Variation in the low-frequency fluctuations (LFFs) of spontaneous brain activity measured by resting-state functional magnetic resonance imaging (RS-fMRI) has been found to reflect individual differences in cognitive measures. In the present study, we aimed to investigate the extent to which initial spontaneous brain activity is related to individual differences in spoken language learning. We acquired RS-fMRI data and subsequently trained participants on a sound-to-word learning paradigm in which they learned to use foreign pitch patterns (from Mandarin Chinese) to signal word meaning. We performed amplitude of spontaneous low-frequency fluctuation (ALFF) analysis, graph theory-based analysis, and independent component analysis (ICA) to identify functional components of the LFFs in the resting-state. First, we examined the ALFF as a regional measure and showed that regional ALFFs in the left superior temporal gyrus were positively correlated with learning performance, whereas ALFFs in the default mode network (DMN) regions were negatively correlated with learning performance. Furthermore, the graph theory-based analysis indicated that the degree and local efficiency of the left superior temporal gyrus were positively correlated with learning performance. Finally, the default mode network and several task-positive resting-state networks (RSNs) were identified via the ICA. The “competition” (i.e., negative correlation) between the DMN and the dorsal attention network was negatively correlated with learning performance. Our results demonstrate that a) spontaneous brain activity can predict future language learning outcome without prior hypotheses (e.g., selection of regions of interest – ROIs) and b) both regional dynamics and network-level interactions in the resting brain can account for individual differences in future spoken language learning success.

Decoupled temporal variability and signal synchronization of spontaneous brain activity in loss of consciousness: An fMRI study in anesthesia

Two aspects of the low frequency fluctuations of spontaneous brain activity have been proposed which reflect the complex and dynamic features of resting-state activity, namely temporal variability and signal synchronization. The relationship between them, especially its role in consciousness, nevertheless remains unclear. Our study examined the temporal variability and signal synchronization of spontaneous brain activity, as well as their relationship during loss of consciousness. We applied an intra-subject design of resting-state functional magnetic resonance imaging (rs-fMRI) in two conditions: during wakefulness, and under anesthesia with clinical unconsciousness. In addition, an independent group of patients with disorders of consciousness (DOC) was included in order to test the reliability of our findings. We observed a global reduction in the temporal variability, local and distant brain signal synchronization for subjects during anesthesia. Importantly, we found a link between temporal variability and both local and distant signal synchronizations during wakefulness: the higher the degree of temporal variability, the higher its intra-regional homogeneity and inter-regional functional connectivity. In contrast, this link was broken down under anesthesia, implying a decoupling between temporal variability and signal synchronization; this decoupling was reproduced in patients with DOC. Our results suggest that there exist some as yet unclear physiological mechanisms of consciousness which “couple” the two mathematically independent measures, temporal variability and signal synchronization of spontaneous brain activity. Our findings not only extend our current knowledge of the neural correlates of anesthetic-induced unconsciousness, but have implications for both computational neural modeling and clinical practice, such as in the diagnosis of loss of consciousness in patients with DOC.

Relating Intrinsic Low-Frequency BOLD Cortical Oscillations to Cognition in Schizophrenia.

The amplitude of low-frequency fluctuations (ALFF) in the blood oxygenation level-dependent (BOLD) signal during resting-state fMRI reflects the magnitude of local low-frequency BOLD oscillations, rather than interregional connectivity. ALFF is of interest to studies of cognition because fluctuations in spontaneous intrinsic brain activity relate to, and possibly even constrain, task-evoked brain responses in healthy people. Lower ALFF has been reported in schizophrenia, but the cognitive correlates of these reductions remain unknown. Here, we assess relationships between ALFF and attention and working memory in order to establish the functional relevance of intrinsic BOLD oscillatory power alterations with respect to specific cognitive impairments in schizophrenia. As part of the multisite FBIRN study, resting-state fMRI data were collected from schizophrenia subjects (SZ; n=168) and healthy controls (HC; n=166). Voxelwise fractional ALFF (fALFF), a normalized ALFF measure, was regressed on neuropsychological measures of sustained attention and working memory in SZ and HC to identify regions showing either common slopes across groups or slope differences between groups (all findings p<0.01 height, p<0.05 family-wise error cluster corrected). Poorer sustained attention was associated with smaller fALFF in the left superior frontal cortex and bilateral temporoparietal junction in both groups, with additional relationships in bilateral posterior parietal, posterior cingulate, dorsal anterior cingulate (ACC), and right dorsolateral prefrontal cortex (DLPFC) evident only in SZ. Poorer working memory was associated with smaller fALFF in bilateral ACC/mPFC, DLPFC, and posterior parietal cortex in both groups. Our findings indicate that smaller amplitudes of low-frequency BOLD oscillations during rest, measured by fALFF, were significantly associated with poorer cognitive performance, sometimes similarly in both groups and sometimes only in SZ, in regions known to subserve sustained attention and working memory. Taken together, these data suggest that the magnitude of resting-state BOLD oscillations shows promise as a biomarker of cognitive function in health and disease.

Mapping intrinsic functional brain changes and repetitive transcranial magnetic stimulation neuromodulation in idiopathic restless legs syndrome: a resting-state functional magnetic resonance imaging study

ObjectiveThe objectives of this study were, first, to explore differences in brain activity between normal people and idiopathic restless legs syndrome (RLS) patients during asymptomatic periods; and, second, to determine whether administering repetitive transcranial magnetic stimulation (rTMS) to specific cortical regions would reverse any observed differences in brain activity and alleviate patient symptoms. MethodsFifteen idiopathic RLS patients (nine drug-naive patients) and 14 gender- and age-matched healthy controls were enrolled. Resting-state functional magnetic resonance imaging was used to measure the amplitude of low-frequency fluctuations (ALFF) in spontaneous brain activity during asymptomatic periods. Seven patients received high-frequency (5 Hz) rTMS directed toward the leg area of the primary motor cortex. Scores on the International Restless Legs Syndrome Study Group (IRLSSG) Rating Scale and ALFF values were measured before and after treatment. ResultsCompared with healthy controls, RLS patients showed lower ALFF in the sensorimotor and visual processing regions, and higher ALFF in the insula, parahippocampal and hippocampal gyri, left posterior parietal areas, and brainstem. These results were largely conserved when only drug-naive patients were considered. After rTMS treatment, ALFF in several sensorimotor and visual regions were significantly elevated and IRLSSG Rating Scale scores decreased, indicating improved RLS symptoms. ConclusionsHigh-frequency rTMS delivered to the leg area of the primary motor cortex may raise functional activity in the sensorimotor and occipital regions, leading to improve symptoms in RLS patients. These results provide novel insight into RLS pathophysiology and suggest a potential mechanism for rTMS therapy in idiopathic RLS patients.

Aberrant functional connectome in neurologically asymptomatic patients with end-stage renal disease.

PurposeThis study aimed to investigate the topological organization of intrinsic functional brain networks in patients with end-stage renal disease (ESRD).Materials and MethodsResting-state functional MRI data were collected from 22 patients with ESRD (16 men, 18–61 years) and 29age- and gender-matched healthy controls (HCs, 19 men, 32–61 years). Whole-brain functional networks were obtained by calculating the interregional correlation of low-frequency fluctuations in spontaneous brain activity among 1,024 parcels that cover the entire cerebrum. Weighted graph-based models were then employed to topologically characterize these networks at different global, modular and nodal levels.ResultsCompared to HCs, the patients exhibited significant disruption in parallel information processing over the whole networks (P< 0.05). The disruption was present in all the functional modules (default mode, executive control, sensorimotor and visual networks) although decreased functional connectivity was observed only within the default mode network. Regional analysis showed that the disease disproportionately weakened nodal efficiency of the default mode components and tended to preferentially affect central or hub-like regions. Intriguingly, the network abnormalities correlated with biochemical hemoglobin and serum calcium levels in the patients. Finally, the functional changes were substantively unchanged after correcting for gray matter atrophy in the patients.ConclusionOur findings provide evidence for the disconnection nature of ESRD’s brain and therefore have important implications for understanding the neuropathologic substrate of the disease from disrupted network organization perspective.

Electrophysiological measures of resting state functional connectivity and their relationship with working memory capacity in childhood.

Functional connectivity is the statistical association of neuronal activity time courses across distinct brain regions, supporting specific cognitive processes. This coordination of activity is likely to be highly important for complex aspects of cognition, such as the communication of fluctuating task goals from higher‐order control regions to lower‐order, functionally specific regions. Some of these functional connections are identifiable even when relevant cognitive tasks are not being performed (i.e. at rest). We used magnetoencephalographic recordings projected into source space to demonstrate that resting state networks in childhood have electrophysiological underpinnings that are evident in the spontaneous fluctuations of oscillatory brain activity. Using the temporal structure of these oscillatory patterns we were able to identify a number of functional resting state networks analogous to those reported in the adult literature. In a second analysis we fused this dynamic temporal information with the spatial information from a functional magnetic resonance imaging analysis of functional connectivity, to demonstrate that inter‐subject variability in these electrophysiological measures of functional connectivity is correlated with individual differences in cognitive ability: the strength of connectivity between a fronto‐parietal network and lower‐level processing areas in inferior temporal cortex was associated with spatial working memory capacity, as measured outside the scanner with educationally relevant standardized assessments. This study represents the first exploration of the electrophysiological mechanisms underpinning resting state functional connectivity in source space in childhood, and the extent to which the strength of particular connections is associated with cognitive ability.

Criticality as a signature of healthy neural systems.

Criticality as a signature of healthy neural systems.

The patterns of fractional amplitude of low-frequency fluctuations in depression patients: The dissociation between temporal regions and fronto-parietal regions

This study surveyed the characteristics of the indicator for the fluctuations of default brain activity, fractional amplitude of low-frequency fluctuations (fALFF), in patients with pure major depressive disorder (MDD). Forty-four patients with MDD and 27 normal controls were enrolled in our study. All the participants received the resting-state functional magnetic resonance imaging (rs-fMRI) scans, which were pre-processed by the REST toolbox (resting-state functional MRI data analysis toolbox). The group-related differences of the fALFF between patients and controls were explored by performing comparisons of the fALFF values obtained from rs-fMRI data. The fALFF outputs of patients and controls were compared with global brain volume, age and gender as covariates. In addition, the correlations between the clinical variables (such as depression severity, anxiety severity, illness duration) and fALFF values were also estimated in each group and across both groups. The patients with MDD had significantly higher fALFF values than the controls, for the left temporal subgyral region. In the contrary, the patients had lower values of fALFF than controls, for the right frontal subcallosal gyrus and right parietal postcentral gyrus. In addition, the fALFF values were negatively correlated with the depression severity in the left temporal subgyral region. MDD patients had dissociable alterations in the fALFF values of the fronto-parietal and temporal regions. The pattern of fALFF alterations might be unique for depression patients with later onset-age of young adult phase.

Large-scale Probabilistic Functional Modes from resting state fMRI

It is well established that it is possible to observe spontaneous, highly structured, fluctuations in human brain activity from functional magnetic resonance imaging (fMRI) when the subject is ‘at rest’. However, characterising this activity in an interpretable manner is still a very open problem.In this paper, we introduce a method for identifying modes of coherent activity from resting state fMRI (rfMRI) data. Our model characterises a mode as the outer product of a spatial map and a time course, constrained by the nature of both the between-subject variation and the effect of the haemodynamic response function. This is presented as a probabilistic generative model within a variational framework that allows Bayesian inference, even on voxelwise rfMRI data. Furthermore, using this approach it becomes possible to infer distinct extended modes that are correlated with each other in space and time, a property which we believe is neuroscientifically desirable.We assess the performance of our model on both simulated data and high quality rfMRI data from the Human Connectome Project, and contrast its properties with those of both spatial and temporal independent component analysis (ICA). We show that our method is able to stably infer sets of modes with complex spatio-temporal interactions and spatial differences between subjects.