Extracting orthogonal subject- and condition-specific signatures from fMRI data using whole-brain effective connectivity

Discovering individual fingerprints in resting-state functional connectivity using deep neural networks

Non-negligible idiosyncrasy due to interindividual differences is an ongoing issue in resting-state functional MRI (rfMRI) analysis. We show that a deep neural network (DNN) can be employed for individual identification by learning important features from the time-varying functional connectivity (FC) of rfMRI in the Human Connectome Project. We employed the trained DNN to identify individuals from an independent dataset acquired at our institution. The results revealed that the DNN could successfully identify 300 individuals with an error rate of 2.9% using 15 s time-window and 870 individuals with an error rate of 6.7%. A trained DNN with nonlinear hidden layers led to the proposal of the "fingerprint of FC" (fpFC) as representative edges of individual FC. The fpFCs for individuals exhibited commonly important and individual-specific edges across time-window lengths (from 5 min to 15 s). Furthermore, the utility of our model for another group of subjects was validated, supporting the feasibility of our technique in the context of transfer learning. In conclusion, our study offers an insight into the discovery of the intrinsic mode of the human brain using whole-brain resting-state FC and DNNs.

Static and dynamic fMRI-derived functional connectomes represent largely similar information

Functional connectivity (FC) of blood oxygen level-dependent (BOLD) fMRI time series can be estimated using methods that differ in sensitivity to the temporal order of time points (static vs. dynamic) and the number of regions considered in estimating a single edge (bivariate vs. multivariate). Previous research suggests that dynamic FC explains variability in FC fluctuations and behavior beyond static FC. Our aim was to systematically compare methods on both dimensions. We compared five FC methods: Pearson's/full correlation (static, bivariate), lagged correlation (dynamic, bivariate), partial correlation (static, multivariate), and multivariate AR model with and without self-connections (dynamic, multivariate). We compared these methods by (i) assessing similarities between FC matrices, (ii) by comparing node centrality measures, and (iii) by comparing the patterns of brain-behavior associations. Although FC estimates did not differ as a function of sensitivity to temporal order, we observed differences between the multivariate and bivariate FC methods. The dynamic FC estimates were highly correlated with the static FC estimates, especially when comparing group-level FC matrices. Similarly, there were high correlations between the patterns of brain-behavior associations obtained using the dynamic and static FC methods. We conclude that the dynamic FC estimates represent information largely similar to that of the static FC.

Group Surrogate Data Generating Models and similarity quantification of multivariate time-series: A resting-state fMRI study

Advancements in non-invasive brain analysis through novel approaches such as big data analytics and in silico simulation are essential for explaining brain function and associated pathologies. In this study, we extend the vector auto-regressive surrogate technique from a single multivariate time-series to group data using a novel Group Surrogate Data Generating Model (GSDGM). This methodology allowed us to generate biologically plausible human brain dynamics representative of a large human resting-state (rs-fMRI) dataset obtained from the Human Connectome Project. Simultaneously, we defined a novel similarity measure, termed the Multivariate Time-series Ensemble Similarity Score (MTESS). MTESS showed high accuracy and f-measure in subject identification, and it can directly compare the similarity between two multivariate time-series. We used MTESS to analyze both human and marmoset rs-fMRI data. Our results showed similarity differences between cortical and subcortical regions. We also conducted MTESS and state transition analysis between single and group surrogate techniques, and confirmed that a group surrogate approach can generate plausible group centroid multivariate time-series. Finally, we used GSDGM and MTESS for the fingerprint analysis of human rs-fMRI data, successfully distinguishing normal and outlier sessions. These new techniques will be useful for clinical applications and in silico simulation.

Estimating dynamic individual coactivation patterns based on densely sampled resting-state fMRI data and utilizing it for better subject identification

As a complex dynamic system, the brain exhibits spatially organized recurring patterns of activity over time. Coactivation patterns (CAPs), which analyzes data from each single frame, have been utilized to detect transient brain activity states recently. However, previous CAP analyses have been conducted at the group level, which might neglect meaningful individual differences. Here, we estimated individual CAP states at both subject- and scan-level based on a densely sampled dataset: Midnight Scan Club. We used differential identifiability, which measures the gap between intra- and inter-subject similarity, to evaluate individual differences. We found individual CAPs at the subject-level achieved the best fingerprinting ability by maintaining high intra-subject similarity and enlarging inter-subject differences, and brain regions of association networks mainly contributed to the identifiability. On the other hand, scan-level CAP states were unstable across scans for the same participant. Expectedly, we found subject-specific CAPs became more reliable and discriminative with more data (i.e., longer duration). As the acquisition time of each participant is limited in practice, our results recommend a data collection strategy that collects more scans with appropriate duration (e.g., 12 ~ 15 min/scan) to obtain more reliable subject-specific CAPs, when total acquisition time is fixed (e.g., 150 min). In summary, this work has constructed reliable subject-specific CAP states with meaningful individual differences, and recommended an appropriate data collection strategy, which can guide subsequent investigations into individualized brain dynamics.

Static and dynamic functional connectomes represent largely similar information

Functional connectivity (FC) of blood-oxygen-level-dependent (BOLD) fMRI time series can be estimated using methods that differ in sensitivity to the temporal order of time points (static vs. dynamic) and the number of regions considered in estimating a single edge (bivariate vs. multivariate). Previous research suggests that dynamic FC explains variability in FC fluctuations and behavior beyond static FC. Our aim was to systematically compare methods on both dimensions. We compared five FC methods: Pearson's/full correlation (static, bivariate), lagged correlation (dynamic, bivariate), partial correlation (static, multivariate) and multivariate AR model with and without self-connections (dynamic, multivariate). We compared these methods by (i) assessing similarities between FC matrices, (ii) by comparing node centrality measures, and (iii) by comparing the patterns of brain-behavior associations. Although FC estimates did not differ as a function of sensitivity to temporal order, we observed differences between the multivariate and bivariate FC methods. The dynamic FC estimates were highly correlated with the static FC estimates, especially when comparing group-level FC matrices. Similarly, there were high correlations between the patterns of brain-behavior associations obtained using the dynamic and static FC methods. We conclude that the dynamic FC estimates represent information largely similar to that of the static FC.

Entropy production of multivariate Ornstein-Uhlenbeck processes correlates with consciousness levels in the human brain

Consciousness is supported by complex patterns of brain activity which are indicative of irreversible nonequilibrium dynamics. While the framework of stochastic thermodynamics has facilitated the understanding of physical systems of this kind, its application to infer the level of consciousness from empirical data remains elusive. We faced this challenge by calculating entropy production in a multivariate Ornstein-Uhlenbeck process fitted to Functional magnetic resonance imaging brain activity recordings. To test this approach, we focused on the transition from wakefulness to deep sleep, revealing a monotonous relationship between entropy production and the level of consciousness. Our results constitute robust signatures of consciousness while also advancing our understanding of the link between consciousness and complexity from the fundamental perspective of statistical physics.

Advancing brain network models to reconcile functional neuroimaging and clinical research

Functional magnetic resonance imaging (fMRI) captures information on brain function beyond the anatomical alterations that are traditionally visually examined by neuroradiologists. However, the fMRI signals are complex in addition to being noisy, so fMRI still faces limitations for clinical applications. Here we review methods that have been proposed as potential solutions so far, namely statistical, biophysical and decoding models, with their strengths and weaknesses. We especially evaluate the ability of these models to directly predict clinical variables from their parameters (predictability) and to extract clinically relevant information regarding biological mechanisms and relevant features for classification and prediction (interpretability). We then provide guidelines for useful applications and pitfalls of such fMRI-based models in a clinical research context, looking beyond the current state of the art. In particular, we argue that the clinical relevance of fMRI calls for a new generation of models for fMRI data, which combine the strengths of both biophysical and decoding models. This leads to reliable and biologically meaningful model parameters, which thus fulfills the need for simultaneous interpretability and predictability. In our view, this synergy is fundamental for the discovery of new pharmacological and interventional targets, as well as the use of models as biomarkers in neurology and psychiatry.

Inferring the dynamical effects of stroke lesions through whole-brain modeling

Understanding the effect of focal lesions (stroke) on brain structure-function traditionally relies on behavioral analyses and correlation with neuroimaging data. Here we use structural disconnection maps from individual lesions to derive a causal mechanistic generative whole-brain model able to explain both functional connectivity alterations and behavioral deficits induced by stroke. As compared to other models that use only the local lesion information, the similarity to the empirical fMRI connectivity increases when the widespread structural disconnection information is considered. The presented model classifies behavioral impairment severity with higher accuracy than other types of information (e.g.: functional connectivity). We assessed topological measures that characterize the functional effects of damage. With the obtained results, we were able to understand how network dynamics change emerge, in a nontrivial way, after a stroke injury of the underlying complex brain system. This type of modeling, including structural disconnection information, helps to deepen our understanding of the underlying mechanisms of stroke lesions.

One session of fMRI-Neurofeedback training on motor imagery modulates whole-brain effective connectivity and dynamical complexity

In the past decade, several studies have shown that Neurofeedback (NFB) by functional magnetic resonance imaging can alter the functional coupling of targeted and non-targeted areas. However, the causal mechanisms underlying these changes remain uncertain. Here, we applied a whole-brain dynamical model to estimate Effective Connectivity (EC) profiles of resting-state data acquired before and immediately after a single-session NFB training for 17 participants who underwent motor imagery NFB training and 16 healthy controls who received sham feedback. Within-group and between-group classification analyses revealed that only for the NFB group it was possible to accurately discriminate between the 2 resting-state sessions. NFB training-related signatures were reflected in a support network of direct connections between areas involved in reward processing and implicit learning, together with regions belonging to the somatomotor, control, attention, and default mode networks, identified through a recursive-feature elimination procedure. By applying a data-driven approach to explore NFB-induced changes in spatiotemporal dynamics, we demonstrated that these regions also showed decreased switching between different brain states (i.e. metastability) only following real NFB training. Overall, our findings contribute to the understanding of NFB impact on the whole brain’s structure and function by shedding light on the direct connections between brain areas affected by NFB training.

Meditation-induced effects on whole-brain structural and effective connectivity

In the past decades, there has been a growing scientific interest in characterizing neural correlates of meditation training. Nonetheless, the mechanisms underlying meditation remain elusive. In the present work, we investigated meditation-related changes in functional dynamics and structural connectivity (SC). For this purpose, we scanned experienced meditators and control (naive) subjects using magnetic resonance imaging (MRI) to acquire structural and functional data during two conditions, resting-state and meditation (focused attention on breathing). In this way, we aimed to characterize and distinguish both short-term and long-term modifications in the brain’s structure and function. First, to analyze the fMRI data, we calculated whole-brain effective connectivity (EC) estimates, relying on a dynamical network model to replicate BOLD signals’ spatio-temporal structure, akin to functional connectivity (FC) with lagged correlations. We compared the estimated EC, FC, and SC links as features to train classifiers to predict behavioral conditions and group identity. Then, we performed a network-based analysis of anatomical connectivity. We demonstrated through a machine-learning approach that EC features were more informative than FC and SC solely. We showed that the most informative EC links that discriminated between meditators and controls involved several large-scale networks mainly within the left hemisphere. Moreover, we found that differences in the functional domain were reflected to a smaller extent in changes at the anatomical level as well. The network-based analysis of anatomical pathways revealed strengthened connectivity for meditators compared to controls between four areas in the left hemisphere belonging to the somatomotor, dorsal attention, subcortical and visual networks. Overall, the results of our whole-brain model-based approach revealed a mechanism underlying meditation by providing causal relationships at the structure-function level.

Improving Functional Connectome Fingerprinting with Degree-Normalization

Background: Functional connectivity quantifies the statistical dependencies between the activity of brain regions, measured using neuroimaging data such as functional magnetic resonance imaging (fMRI) blood-oxygenation-level dependent time series. The network representation of functional connectivity, called a functional connectome (FC), has been shown to contain an individual fingerprint allowing participants identification across consecutive testing sessions. Recently, researchers have focused on the extraction of these fingerprints, with potential applications in personalized medicine. Materials and Methods: In this study, we show that a mathematical operation denominated degree-normalization can improve the extraction of FC fingerprints. Degree-normalization has the effect of reducing the excessive influence of strongly connected brain areas in the whole-brain network. We adopt the differential identifiability framework and apply it to both original and degree-normalized FCs of 409 individuals from the Human Connectome Project, in resting-state and 7 fMRI tasks. Results: Our results indicate that degree-normalization systematically improves three fingerprinting metrics, namely differential identifiability, identification rate, and matching rate. Moreover, the results related to the matching rate metric suggest that individual fingerprints are embedded in a low-dimensional space. Discussion: The results suggest that low-dimensional functional fingerprints lie in part in weakly connected subnetworks of the brain and that degree-normalization helps uncovering them. This work introduces a simple mathematical operation that could lead to significant improvements in future FC fingerprinting studies.

A machine learning approach to characterize sequential movement-related states in premotor and motor cortices

Non-human primate (NHP) movement kinematics have been decoded from spikes and local field potentials (LFPs) recorded during motor tasks. However, the potential of LFPs to provide network-like characterizations of neural dynamics during planning and execution of sequential movements requires further exploration. Is the aggregate nature of LFPs suitable to construct informative brain state descriptors of movement preparation and execution? To investigate this, we developed a framework to process LFPs based on machine learning classifiers and analyzed LFP from a primate, implanted with several microelectrode arrays covering the premotor cortex in both hemispheres and the primary motor cortex on one side. The primate performed a reach and grasp task, consisting of five consecutive states, starting from rest until a rewarding target (food) was attained. We use this five-state task to characterize brain activity and connectivity within eight frequency bands, using spectral amplitude and pair-wise correlations across electrodes as features. Our results show that we could best distinguish all five movement-related states using the highest frequency band (200-500Hz), yielding an 87% accuracy with spectral amplitude, and 60% with pair-wise electrode correlation. Further analyses characterized each movement-related state, showing differential neuronal population activity at above-gamma frequencies during the various stages of movement. In summary, our results show that the concerted use of novel machine-learning techniques with coarse grained queue broad signals such as LFPs may be successfully used to track and decode fine reach and grasp movement aspects across several brain regions.

Effective connectivity extracts clinically relevant prognostic information from resting state activity in stroke

Recent resting-state functional MRI studies in stroke patients have identified two robust biomarkers of acute brain dysfunction: a reduction of inter-hemispheric functional connectivity between homotopic regions of the same network, and an abnormal increase of ipsi-lesional functional connectivity between task-negative and task-positive resting-state networks. Whole-brain computational modelling studies, at the individual subject level, using undirected effective connectivity derived from empirically measured functional connectivity, have shown a reduction of measures of integration and segregation in stroke as compared to healthy brains. Here we employ a novel method, first, to infer whole-brain directional effective connectivity from zero-lagged and lagged covariance matrices, then, to compare it to empirically measured functional connectivity for predicting stroke versus healthy status, and patient performance (zero, one, multiple deficits) across neuropsychological tests. We also investigated the accuracy of functional connectivity versus model effective connectivity in predicting the long-term outcome from acute measures. Both functional and effective connectivity predicted healthy from stroke individuals significantly better than the chance-level; however, accuracy for the effective connectivity was significantly higher than for functional connectivity at 1- to 2-week, 3-month and 1-year post-stroke. Predictive functional connections mainly included those reported in previous studies (within-network inter-hemispheric and between task-positive and -negative networks intra-hemispherically). Predictive effective connections included additional between-network links. Effective connectivity was a better predictor than functional connectivity of the number of behavioural domains in which patients suffered deficits, both at 2-week and 1-year post-onset of stroke. Interestingly, patient deficits at 1-year time-point were better predicted by effective connectivity values at 2 weeks rather than at 1-year time-point. Our results thus demonstrate that the second-order statistics of functional MRI resting-state activity at an early stage of stroke, derived from a whole-brain effective connectivity, estimated in a model fitted to reproduce the propagation of neuronal activity, has pertinent information for clinical prognosis.

Optimizing differential identifiability improves connectome predictive modeling of cognitive deficits from functional connectivity in Alzheimer's disease

Functional connectivity, as estimated using resting state functional MRI, has shown potential in bridging the gap between pathophysiology and cognition. However, clinical use of functional connectivity biomarkers is impeded by unreliable estimates of individual functional connectomes and lack of generalizability of models predicting cognitive outcomes from connectivity. To address these issues, we combine the frameworks of connectome predictive modeling and differential identifiability. Using the combined framework, we show that enhancing the individual fingerprint of resting state functional connectomes leads to robust identification of functional networks associated to cognitive outcomes and also improves prediction of cognitive outcomes from functional connectomes. Using a comprehensive spectrum of cognitive outcomes associated to Alzheimer's disease (AD), we identify and characterize functional networks associated to specific cognitive deficits exhibited in AD. This combined framework is an important step in making individual level predictions of cognition from resting state functional connectomes and in understanding the relationship between cognition and connectivity.

Beyond the disconnectivity hypothesis of schizophrenia

To go beyond the disconnectivity hypothesis of schizophrenia, directed (effective) connectivity was measured between 94 brain regions, to provide evidence on the source of the changes in schizophrenia and a mechanistic model. Effective connectivity (EC) was measured in 180 participants with schizophrenia and 208 controls. For the significantly different effective connectivities in schizophrenia, on average the forward (stronger) effective connectivities were smaller, whereas the backward connectivities tended to be larger. Further, higher EC in schizophrenia was found from the precuneus and posterior cingulate cortex (PCC) to areas such as the parahippocampal, hippocampal, temporal, fusiform, and occipital cortices. These are backward effective connectivities and were positively correlated with the positive symptoms of schizophrenia. Lower effective connectivities were found from temporal and other regions and were negatively correlated with the symptoms, especially the negative and general symptoms. Further, a signal variance parameter was increased for areas that included the parahippocampal gyrus and hippocampus, consistent with the hypothesis that hippocampal overactivity is involved in schizophrenia. This investigation goes beyond the disconnectivity hypothesis by drawing attention to differences in schizophrenia between backprojections and forward connections, with the backward connections from the precuneus and PCC implicated in memory stronger in schizophrenia.

Geodesic Distance on Optimally Regularized Functional Connectomes Uncovers Individual Fingerprints

Background: Functional connectomes (FCs) have been shown to provide a reproducible individual fingerprint, which has opened the possibility of personalized medicine for neuro/psychiatric disorders. Thus, developing accurate ways to compare FCs is essential to establish associations with behavior and/or cognition at the individual level. Methods: Canonically, FCs are compared using Pearson's correlation coefficient of the entire functional connectivity profiles. Recently, it has been proposed that the use of geodesic distance is a more accurate way of comparing FCs, one which reflects the underlying non-Euclidean geometry of the data. Computing geodesic distance requires FCs to be positive-definite and hence invertible matrices. As this requirement depends on the functional magnetic resonance imaging scanning length and the parcellation used, it is not always attainable and sometimes a regularization procedure is required. Results: In the present work, we show that regularization is not only an algebraic operation for making FCs invertible, but also that an optimal magnitude of regularization leads to systematically higher fingerprints. We also show evidence that optimal regularization is data set-dependent and varies as a function of condition, parcellation, scanning length, and the number of frames used to compute the FCs. Discussion: We demonstrate that a universally fixed regularization does not fully uncover the potential of geodesic distance on individual fingerprinting and indeed could severely diminish it. Thus, an optimal regularization must be estimated on each data set to uncover the most differentiable across-subject and reproducible within-subject geodesic distances between FCs. The resulting pairwise geodesic distances at the optimal regularization level constitute a very reliable quantification of differences between subjects.

Brain States and Transitions: Insights from Computational Neuroscience

Within the field of computational neuroscience there are great expectations of finding new ways to rebalance the complex dynamic system of the human brain through controlled pharmacological or electromagnetic perturbation. Yet many obstacles remain between the ability to accurately predict how and where best to perturb to force a transition from one brain state to another. The foremost challenge is a commonly agreed definition of a given brain state. Recent progress in computational neuroscience has made it possible to robustly define brain states and force transitions between them. Here, we review the state of the art and propose a framework for determining the functional hierarchical organization describing any given brain state. We describe the latest advances in creating sophisticated whole-brain computational models with interacting neuronal and neurotransmitter systems that can be studied fully in silico to predict and design novel pharmacological and electromagnetic interventions to rebalance them in disease.

Computational Models in Electroencephalography

Computational models lie at the intersection of basic neuroscience and healthcare applications because they allow researchers to test hypotheses in silico and predict the outcome of experiments and interactions that are very hard to test in reality. Yet, what is meant by “computational model” is understood in many different ways by researchers in different fields of neuroscience and psychology, hindering communication and collaboration. In this review, we point out the state of the art of computational modeling in Electroencephalography (EEG) and outline how these models can be used to integrate findings from electrophysiology, network-level models, and behavior. On the one hand, computational models serve to investigate the mechanisms that generate brain activity, for example measured with EEG, such as the transient emergence of oscillations at different frequency bands and/or with different spatial topographies. On the other hand, computational models serve to design experiments and test hypotheses in silico. The final purpose of computational models of EEG is to obtain a comprehensive understanding of the mechanisms that underlie the EEG signal. This is crucial for an accurate interpretation of EEG measurements that may ultimately serve in the development of novel clinical applications.

Resting-State Co-activation Patterns as Promising Candidates for Prediction of Alzheimer's Disease in Aged Mice

Alzheimer’s disease (AD), a neurodegenerative disorder marked by accumulation of extracellular amyloid-β (Aβ) plaques leads to progressive loss of memory and cognitive function. Resting-state fMRI (RS-fMRI) studies have provided links between these two observations in terms of disruption of default mode and task-positive resting-state networks (RSNs). Important insights underlying these disruptions were recently obtained by investigating dynamic fluctuations in RS-fMRI signals in old TG2576 mice (a mouse model of amyloidosis) using a set of quasi-periodic patterns (QPP). QPPs represent repeating spatiotemporal patterns of neural activity of predefined temporal length. In this article, we used an alternative methodology of co-activation patterns (CAPs) that represent instantaneous and transient brain configurations that are likely contributors to the emergence of commonly observed RSNs and QPPs. We followed a recently published approach for obtaining CAPs that divided all time frames, instead of those corresponding to supra-threshold activations of a seed region as done traditionally, to extract CAPs from RS-fMRI recordings in 10 TG2576 female mice and eight wild type littermates at 18 months of age. Subsequently, we matched the CAPs from the two groups using the Hungarian method and compared the temporal (duration, occurrence rate) and the spatial (lateralization of significantly co-activated and co-deactivated voxels) properties of matched CAPs. We found robust differences in the spatial components of matched CAPs. Finally, we used supervised learning to train a classifier using either the temporal or the spatial component of CAPs to distinguish the transgenic mice from the WT. We found that while duration and occurrence rates of all CAPs performed the classification with significantly higher accuracy than the chance-level, blood oxygen level-dependent (BOLD) signals of significantly activated voxels from individual CAPs turned out to be a significantly better predictive feature demonstrating a near-perfect classification accuracy. Our results demonstrate resting-state co-activation patterns are a promising candidate in the development of a diagnostic, and potentially, prognostic RS-fMRI biomarker of AD.

Diagnosis of major depressive disorder using whole-brain effective connectivity networks derived from resting-state functional MRI

OBJECTIVE

It is important to improve identification accuracy for possible early intervention of major depressive disorder (MDD). Recently, effective connectivity (EC), defined as the directed influence of spatially distant brain regions on each other, has been used to find the dysfunctional organization of brain networks in MDD. However, little is known about the ability of whole-brain resting-state EC features in identification of MDD. Here, we employed EC by whole-brain analysis to perform MDD diagnosis.

APPROACH

In this study, we proposed a high-order EC network capturing high-level relationship among multiple brain regions to discriminate 57 patients with MDD from 60 normal controls (NC). In high-order EC networks and traditional low-order EC networks, we utilized the network properties and connection strength for classification. Meanwhile, the support vector machine (SVM) was employed for model training. Generalization of the results was supported by 10-fold cross-validation.

MAIN RESULTS

The classification results showed that the high-order EC network performed better than the low-order EC network in diagnosing MDD, and the integration of these two networks yielded the best classification precision with 95% accuracy, 98.83% sensitivity, and 91% specificity. Furthermore, we found that the abnormal connections of high-order EC in MDD patients involved multiple widely concerned functional subnets, particularly the default mode network and the cerebellar network.

SIGNIFICANCE

The current study indicates whole-brain EC networks, measured by our high-order method, may be promising biomarkers for clinical diagnosis of MDD, and the complementary between high-order and low-order EC will better guide patients to get early interventions as well as treatments.

GEFF: Graph embedding for functional fingerprinting

It has been well established that Functional Connectomes (FCs), as estimated from functional MRI (fMRI) data, have an individual fingerprint that can be used to identify an individual from a population (subject-identification). Although identification rate is high when using resting-state FCs, other tasks show moderate to low values. Furthermore, identification rate is task-dependent, and is low when distinct cognitive states, as captured by different fMRI tasks, are compared. Here we propose an embedding framework, GEFF (Graph Embedding for Functional Fingerprinting), based on group-level decomposition of FCs into eigenvectors. GEFF creates an eigenspace representation of a group of subjects using one or more task FCs (Learning Stage). In the Identification Stage, we compare new instances of FCs from the Learning subjects within this eigenspace (validation dataset). The validation dataset contains FCs either from the same tasks as the Learning dataset or from the remaining tasks that were not included in Learning. Assessment of validation FCs within the eigenspace results in significantly increased subject-identification rates for all fMRI tasks tested and potentially task-independent fingerprinting process. It is noteworthy that combining resting-state with one fMRI task for GEFF Learning Stage covers most of the cognitive space for subject identification. Thus, while designing an experiment, one could choose a task fMRI to ask a specific question and combine it with resting-state fMRI to extract maximum subject differentiability using GEFF. In addition to subject-identification, GEFF was also used for identification of cognitive states, i.e. to identify the task associated to a given FC, regardless of the subject being already in the Learning dataset or not (subject-independent task-identification). In addition, we also show that eigenvectors from the Learning Stage can be characterized as task- and subject-dominant, subject-dominant or neither, using two-way ANOVA of their corresponding loadings, providing a deeper insight into the extent of variance in functional connectivity across individuals and cognitive states.

Uncovering differential identifiability in network properties of human brain functional connectomes

The identifiability framework (𝕀f) has been shown to improve differential identifiability (reliability across-sessions and -sites, and differentiability across-subjects) of functional connectomes for a variety of fMRI tasks. But having a robust single session/subject functional connectome is just the starting point to subsequently assess network properties for characterizing properties of integration, segregation, and communicability, among others. Naturally, one wonders whether uncovering identifiability at the connectome level also uncovers identifiability on the derived network properties. This also raises the question of where to apply the 𝕀f framework: on the connectivity data or directly on each network measurement? Our work answers these questions by exploring the differential identifiability profiles of network measures when 𝕀f is applied (a) on the functional connectomes, and (b) directly on derived network measurements. Results show that improving across-session reliability of functional connectomes (FCs) also improves reliability of derived network measures. We also find that, for specific network properties, application of 𝕀f directly on network properties is more effective. Finally, we discover that applying the framework, either way, increases task sensitivity of network properties. At a time when the neuroscientific community is focused on subject-level inferences, this framework is able to uncover FC fingerprints, which propagate to derived network properties.

Stable biomarker identification for predicting schizophrenia in the human connectome

Schizophrenia, as a psychiatric disorder, has recognized brain alterations both at the structural and at the functional magnetic resonance imaging level. The developing field of connectomics has attracted much attention as it allows researchers to take advantage of powerful tools of network analysis in order to study structural and functional connectivity abnormalities in schizophrenia. Many methods have been proposed to identify biomarkers in schizophrenia, focusing mainly on improving the classification performance or performing statistical comparisons between groups. However, the stability of biomarkers selection has been for long overlooked in the connectomics field. In this study, we follow a machine learning approach where the identification of biomarkers is addressed as a feature selection problem for a classification task. We perform a recursive feature elimination and support vector machines (RFE-SVM) approach to identify the most meaningful biomarkers from the structural, functional, and multi-modal connectomes of healthy controls and patients. Furthermore, the stability of the retrieved biomarkers is assessed across different subsamplings of the dataset, allowing us to identify the affected core of the pathology. Considering our technique altogether, it demonstrates a principled way to achieve both accurate and stable biomarkers while highlighting the importance of multi-modal approaches to brain pathology as they tend to reveal complementary information.

Accurate prediction of individual subject identity and task, but not autism diagnosis, from functional connectomes

Despite enthusiasm about the potential for using fMRI-based functional connectomes in the development of biomarkers for autism spectrum disorder (ASD), the literature is full of negative findings-failures to distinguish ASD functional connectomes from those of typically developing controls (TD)-and positive findings that are inconsistent across studies. Here, we report on a new study designed to either better differentiate ASD from TD functional connectomes-or, alternatively, to refine our understanding of the factors underlying the current state of affairs. We scanned individuals with ASD and controls both at rest and while watching videos with social content. Using multiband fMRI across repeat sessions, we improved both data quantity and scanning duration by collecting up to 2 hr of data per individual. This is about 50 times the typical number of temporal samples per individual in ASD fcMRI studies. We obtained functional connectomes that were discriminable, allowing for near-perfect individual identification regardless of diagnosis, and equally reliable in both groups. However, contrary to what one might expect, we did not consistently or robustly observe in the ASD group either reductions in similarity to TD functional connectivity (FC) patterns or shared atypical FC patterns. Accordingly, FC-based predictions of diagnosis group achieved accuracy levels around chance. However, using the same approaches to predict scan type (rest vs. video) achieved near-perfect accuracy. Our findings suggest that neither the limitations of resting state as a "task," data resolution, data quantity, or scan duration can be considered solely responsible for failures to differentiate ASD from TD functional connectomes.

Dynamic coupling of whole-brain neuronal and neurotransmitter systems

Remarkable progress has come from whole-brain models linking anatomy and function. Paradoxically, it is not clear how a neuronal dynamical system running in the fixed human anatomical connectome can give rise to the rich changes in the functional repertoire associated with human brain function, which is impossible to explain through long-term plasticity. Neuromodulation evolved to allow for such flexibility by dynamically updating the effectivity of the fixed anatomical connectivity. Here, we introduce a theoretical framework modeling the dynamical mutual coupling between the neuronal and neurotransmitter systems. We demonstrate that this framework is crucial to advance our understanding of whole-brain dynamics by bidirectional coupling of the two systems through combining multimodal neuroimaging data (diffusion magnetic resonance imaging [dMRI], functional magnetic resonance imaging [fMRI], and positron electron tomography [PET]) to explain the functional effects of specific serotoninergic receptor (5-HT_2AR) stimulation with psilocybin in healthy humans. This advance provides an understanding of why psilocybin is showing considerable promise as a therapeutic intervention for neuropsychiatric disorders including depression, anxiety, and addiction. Overall, these insights demonstrate that the whole-brain mutual coupling between the neuronal and the neurotransmission systems is essential for understanding the remarkable flexibility of human brain function despite having to rely on fixed anatomical connectivity.

Model-based whole-brain effective connectivity to study distributed cognition in health and disease

Neuroimaging techniques are now widely used to study human cognition. The functional associations between brain areas have become a standard proxy to describe how cognitive processes are distributed across the brain network. Among the many analysis tools available, dynamic models of brain activity have been developed to overcome the limitations of original connectivity measures such as functional connectivity. This goes in line with the many efforts devoted to the assessment of directional interactions between brain areas from the observed neuroimaging activity. This opinion article provides an overview of our model-based whole-brain effective connectivity to analyze fMRI data, while discussing the pros and cons of our approach with respect to other established approaches. Our framework relies on the multivariate Ornstein-Uhlenbeck (MOU) process and is thus referred to as MOU-EC. Once tuned, the model provides a directed connectivity estimate that reflects the dynamical state of BOLD activity, which can be used to explore cognition. We illustrate this approach using two applications on task-evoked fMRI data. First, as a connectivity measure, MOU-EC can be used to extract biomarkers for task-specific brain coordination, understood as the patterns of areas exchanging information. The multivariate nature of connectivity measures raises several challenges for whole-brain analysis, for which machine-learning tools present some advantages over statistical testing. Second, we show how to interpret changes in MOU-EC connections in a collective and model-based manner, bridging with network analysis. Our framework provides a comprehensive set of tools that open exciting perspectives to study distributed cognition, as well as neuropathologies.

Network analysis of whole-brain fMRI dynamics: A new framework based on dynamic communicability

Neuroimaging techniques such as MRI have been widely used to explore the associations between brain areas. Structural connectivity (SC) captures the anatomical pathways across the brain and functional connectivity (FC) measures the correlation between the activity of brain regions. These connectivity measures have been much studied using network theory in order to uncover the distributed organization of brain structures, in particular FC for task-specific brain communication. However, the application of network theory to study FC matrices is often "static" despite the dynamic nature of time series obtained from fMRI. The present study aims to overcome this limitation by introducing a network-oriented analysis applied to whole-brain effective connectivity (EC) useful to interpret the brain dynamics. Technically, we tune a multivariate Ornstein-Uhlenbeck (MOU) process to reproduce the statistics of the whole-brain resting-state fMRI signals, which provides estimates for MOU-EC as well as input properties (similar to local excitabilities). The network analysis is then based on the Green function (or network impulse response) that describes the interactions between nodes across time for the estimated dynamics. This model-based approach provides time-dependent graph-like descriptor, named communicability, that characterize the roles that either nodes or connections play in the propagation of activity within the network. They can be used at both global and local levels, and also enables the comparison of estimates from real data with surrogates (e.g. random network or ring lattice). In contrast to classical graph approaches to study SC or FC, our framework stresses the importance of taking the temporal aspect of fMRI signals into account. Our results show a merging of functional communities over time, moving from segregated to global integration of the network activity. Our formalism sets a solid ground for the analysis and interpretation of fMRI data, including task-evoked activity.

A new computational approach to estimate whole-brain effective connectivity from functional and structural MRI, applied to language development

Recently introduced effective connectivity methods allow for the in-vivo investigation of large-scale functional interactions between brain regions. However, dynamic causal modeling, the most widely used technique to date, typically captures only a few predefined regions of interest. In this study, we present an alternative computational approach to infer effective connectivity within the entire connectome and show its performance on a developmental cohort with emerging language capacities. The novel approach provides new opportunities to quantify effective connectivity changes in the human brain.