Astrocytes as a mechanism for contextually-guided network dynamics and function.

Astrocytes are a ubiquitous and enigmatic type of non-neuronal cell and are found in the brain of all vertebrates. While traditionally viewed as being supportive of neurons, it is increasingly recognized that astrocytes play a more direct and active role in brain function and neural computation. On account of their sensitivity to a host of physiological covariates and ability to modulate neuronal activity and connectivity on slower time scales, astrocytes may be particularly well poised to modulate the dynamics of neural circuits in functionally salient ways. In the current paper, we seek to capture these features via actionable abstractions within computational models of neuron-astrocyte interaction. Specifically, we engage how nested feedback loops of neuron-astrocyte interaction, acting over separated time-scales, may endow astrocytes with the capability to enable learning in context-dependent settings, where fluctuations in task parameters may occur much more slowly than within-task requirements. We pose a general model of neuron-synapse-astrocyte interaction and use formal analysis to characterize how astrocytic modulation may constitute a form of meta-plasticity, altering the ways in which synapses and neurons adapt as a function of time. We then embed this model in a bandit-based reinforcement learning task environment, and show how the presence of time-scale separated astrocytic modulation enables learning over multiple fluctuating contexts. Indeed, these networks learn far more reliably compared to dynamically homogeneous networks and conventional non-network-based bandit algorithms. Our results fuel the notion that neuron-astrocyte interactions in the brain benefit learning over different time-scales and the conveyance of task-relevant contextual information onto circuit dynamics.

Heterogeneous Forgetting Rates and Greedy Allocation in Slot-Based Memory Networks Promotes Signal Retention.

A key question in the neuroscience of memory encoding pertains to the mechanisms by which afferent stimuli are allocated within memory networks. This issue is especially pronounced in the domain of working memory, where capacity is finite. Presumably the brain must embed some "policy" by which to allocate these mnemonic resources in an online manner in order to maximally represent and store afferent information for as long as possible and without interference from subsequent stimuli. Here, we engage this question through a top-down theoretical modeling framework. We formally optimize a gating mechanism that projects afferent stimuli onto a finite number of memory slots within a recurrent network architecture. In the absence of external input, the activity in each slot attenuates over time (i.e., a process of gradual forgetting). It turns out that the optimal gating policy consists of a direct projection from sensory activity to memory slots, alongside an activity-dependent lateral inhibition. Interestingly, allocating resources myopically (greedily with respect to the current stimulus) leads to efficient utilization of slots over time. In other words, later-arriving stimuli are distributed across slots in such a way that the network state is minimally shifted and so prior signals are minimally "overwritten." Further, networks with heterogeneity in the timescales of their forgetting rates retain stimuli better than those that are more homogeneous. Our results suggest how online, recurrent networks working on temporally localized objectives without high-level supervision can nonetheless implement efficient allocation of memory resources over time.

Computational modelling in disorders of consciousness: Closing the gap towards personalised models for restoring consciousness.

Disorders of consciousness are complex conditions characterised by persistent loss of responsiveness due to brain injury. They present diagnostic challenges and limited options for treatment, and highlight the urgent need for a more thorough understanding of how human consciousness arises from coordinated neural activity. The increasing availability of multimodal neuroimaging data has given rise to a wide range of clinically- and scientifically-motivated modelling efforts, seeking to improve data-driven stratification of patients, to identify causal mechanisms for patient pathophysiology and loss of consciousness more broadly, and to develop simulations as a means of testing in silico potential treatment avenues to restore consciousness. As a dedicated Working Group of clinicians and neuroscientists of the international Curing Coma Campaign, here we provide our framework and vision to understand the diverse statistical and generative computational modelling approaches that are being employed in this fast-growing field. We identify the gaps that exist between the current state-of-the-art in statistical and biophysical computational modelling in human neuroscience, and the aspirational goal of a mature field of modelling disorders of consciousness; which might drive improved treatments and outcomes in the clinic. Finally, we make several recommendations for how the field as a whole can work together to address these challenges.

Enhancing task fMRI preprocessing via individualized model-based filtering of intrinsic activity dynamics.

Brain responses recorded during fMRI are thought to reflect both rapid, stimulus-evoked activity and the propagation of spontaneous activity through brain networks. In the current work, we describe a method to improve the estimation of task-evoked brain activity by first "filtering-out the intrinsic propagation of pre-event activity from the BOLD signal. We do so using Mesoscale Individualized NeuroDynamic (MINDy; Singh et al. 2020b) models built from individualized resting-state data to subtract the propagation of spontaneous activity from the task-fMRI signal (MINDy-based Filtering). After filtering, time-series are analyzed using conventional techniques. Results demonstrate that this simple operation significantly improves the statistical power and temporal precision of estimated group-level effects. Moreover, use of MINDy-based filtering increased the similarity of neural activation profiles and prediction accuracy of individual differences in behavior across tasks measuring the same construct (cognitive control). Thus, by subtracting the propagation of previous activity, we obtain better estimates of task-related neural effects.

Slow manifolds within network dynamics encode working memory efficiently and robustly.

Working memory is a cognitive function involving the storage and manipulation of latent information over brief intervals of time, thus making it crucial for context-dependent computation. Here, we use a top-down modeling approach to examine network-level mechanisms of working memory, an enigmatic issue and central topic of study in neuroscience. We optimize thousands of recurrent rate-based neural networks on a working memory task and then perform dynamical systems analysis on the ensuing optimized networks, wherein we find that four distinct dynamical mechanisms can emerge. In particular, we show the prevalence of a mechanism in which memories are encoded along slow stable manifolds in the network state space, leading to a phasic neuronal activation profile during memory periods. In contrast to mechanisms in which memories are directly encoded at stable attractors, these networks naturally forget stimuli over time. Despite this seeming functional disadvantage, they are more efficient in terms of how they leverage their attractor landscape and paradoxically, are considerably more robust to noise. Our results provide new hypotheses regarding how working memory function may be encoded within the dynamics of neural circuits.

Developing control-theoretic objectives for large-scale brain dynamics and cognitive enhancement.

The development of technologies for brain stimulation provides a means for scientists and clinicians to directly actuate the brain and nervous system. Brain stimulation has shown intriguing potential in terms of modifying particular symptom clusters in patients and behavioral characteristics of subjects. The stage is thus set for optimization of these techniques and the pursuit of more nuanced stimulation objectives, including the modification of complex cognitive functions such as memory and attention. Control theory and engineering will play a key role in the development of these methods, guiding computational and algorithmic strategies for stimulation. In particular, realizing this goal will require new development of frameworks that allow for controlling not only brain activity, but also latent dynamics that underlie neural computation and information processing. In the current opinion, we review recent progress in brain stimulation and outline challenges and potential research pathways associated with exogenous control of cognitive function.

Neural Circuit Dynamics for Sensory Detection.

We consider the question of how sensory networks enable the detection of sensory stimuli in a combinatorial coding space. We are specifically interested in the olfactory system, wherein recent experimental studies have reported the existence of rich, enigmatic response patterns associated with stimulus onset and offset. This study aims to identify the functional relevance of such response patterns (i.e., what benefits does such neural activity provide in the context of detecting stimuli in a natural environment). We study this problem through the lens of normative, optimization-based modeling. Here, we define the notion of a low-dimensional latent representation of stimulus identity, which is generated through action of the sensory network. The objective of our optimization framework is to ensure high-fidelity tracking of a nominal representation in this latent space in an energy-efficient manner. It turns out that the optimal motifs emerging from this framework possess morphologic similarity with prototypical onset and offset responses observed in vivo in locusts (Schistocerca americana) of either sex. Furthermore, this objective can be exactly achieved by a network with reciprocal excitatory-inhibitory competitive dynamics, similar to interactions between projection neurons and local neurons in the early olfactory system of insects. The derived model also makes several predictions regarding maintenance of robust latent representations in the presence of confounding background information and trade-offs between the energy of sensory activity and resultant behavioral measures such as speed and accuracy of stimulus detection.SIGNIFICANCE STATEMENT A key area of study in olfactory coding involves understanding the transformation from high-dimensional sensory stimulus to low-dimensional decoded representation. Here, we examine not only the dimensionality reduction of this mapping but also its temporal dynamics, with specific focus on stimuli that are temporally continuous. Through optimization-based synthesis, we examine how sensory networks can track representations without prior assumption of discrete trial structure. We show that such tracking can be achieved by canonical network architectures and dynamics, and that the resulting responses resemble observations from neurons in the insect olfactory system. Thus, our results provide hypotheses regarding the functional role of olfactory circuit activity at both single neuronal and population scales.

Estimation and validation of individualized dynamic brain models with resting state fMRI.

A key challenge for neuroscience is to develop generative, causal models of the human nervous system in an individualized, data-driven manner. Previous initiatives have either constructed biologically-plausible models that are not constrained by individual-level human brain activity or used data-driven statistical characterizations of individuals that are not mechanistic. We aim to bridge this gap through the development of a new modeling approach termed Mesoscale Individualized Neurodynamic (MINDy) modeling, wherein we fit nonlinear dynamical systems models directly to human brain imaging data. The MINDy framework is able to produce these data-driven network models for hundreds to thousands of interacting brain regions in just 1-3 â€‹min per subject. We demonstrate that the models are valid, reliable, and robust. We show that MINDy models are predictive of individualized patterns of resting-state brain dynamical activity. Furthermore, MINDy is better able to uncover the mechanisms underlying individual differences in resting state activity than functional connectivity methods.

Biophysically interpretable inference of single neuron dynamics.

Identification of key ionic channel contributors to the overall dynamics of a neuron is an important problem in experimental neuroscience. Such a problem is challenging since even in the best cases, identification relies on noisy recordings of membrane potential only, and strict inversion to the constituent channel dynamics is mathematically ill-posed. In this work, we develop a biophysically interpretable, learning-based strategy for data-driven inference of neuronal dynamics. In particular, we propose two optimization frameworks to learn and approximate neural dynamics from an observed voltage trajectory. In both the proposed strategies, the membrane potential dynamics are approximated as a weighted sum of ionic currents. In the first strategy, the ionic currents are represented using voltage dependent channel conductances and membrane potential in a parametric form, while in the second strategy, the currents are represented as a linear combination of generic basis functions. A library of channel activation/inactivation and time-constant curves describing prototypical channel kinetics are used to provide estimates of the channel variables to approximate the ionic currents. Finally, a linear optimization problem is solved to infer the weights/scaling variables in the membrane-potential dynamics. In the first strategy, the weights can be used to recover the channel conductances, and the reversal potentials while in the second strategy, using the estimated weights, active channels can be inferred and the trajectory of the gating variables are recovered, allowing for biophysically salient inference. Our results suggest that the complex nonlinear behavior of the neural dynamics over a range of temporal scales can be efficiently inferred in a data-driven manner from noisy membrane potential recordings.

Multiple Timescale Online Learning Rules for Information Maximization with Energetic Constraints.

A key aspect of the neural coding problem is understanding how representations of afferent stimuli are built through the dynamics of learning and adaptation within neural networks. The infomax paradigm is built on the premise that such learning attempts to maximize the mutual information between input stimuli and neural activities. In this letter, we tackle the problem of such information-based neural coding with an eye toward two conceptual hurdles. Specifically, we examine and then show how this form of coding can be achieved with online input processing. Our framework thus obviates the biological incompatibility of optimization methods that rely on global network awareness and batch processing of sensory signals. Central to our result is the use of variational bounds as a surrogate objective function, an established technique that has not previously been shown to yield online policies. We obtain learning dynamics for both linear-continuous and discrete spiking neural encoding models under the umbrella of linear gaussian decoders. This result is enabled by approximating certain information quantities in terms of neuronal activity via pairwise feedback mechanisms. Furthermore, we tackle the problem of how such learning dynamics can be realized with strict energetic constraints. We show that endowing networks with auxiliary variables that evolve on a slower timescale can allow for the realization of saddle-point optimization within the neural dynamics, leading to neural codes with favorable properties in terms of both information and energy.

Spiking networks as efficient distributed controllers.

In the brain, networks of neurons produce activity that is decoded into perceptions and actions. How the dynamics of neural networks support this decoding is a major scientific question. That is, while we understand the basic mechanisms by which neurons produce activity in the form of spikes, whether these dynamics reflect an overlying functional objective is not understood. In this paper, we examine neuronal dynamics from a first-principles control-theoretic viewpoint. Specifically, we postulate an objective wherein neuronal spiking activity is decoded into a control signal that subsequently drives a linear system. Then, using a recently proposed principle from theoretical neuroscience, we optimize the production of spikes so that the linear system in question achieves reference tracking. It turns out that such optimization leads to a recurrent network architecture wherein each neuron possess integrative dynamics. The network amounts to an efficient, distributed event-based controller where each neuron (node) produces a spike if doing so improves tracking performance. Moreover, the dynamics provide inherent robustness properties, so that if some neurons fail, others will compensate by increasing their activity so that the tracking objective is met.

Recurrent Information Optimization with Local, Metaplastic Synaptic Dynamics.

We consider the problem of optimizing information-theoretic quantities in recurrent networks via synaptic learning. In contrast to feedforward networks, the recurrence presents a key challenge insofar as an optimal learning rule must aggregate the joint distribution of the whole network. This challenge, in particular, makes a local policy (i.e., one that depends on only pairwise interactions) difficult. Here, we report a local metaplastic learning rule that performs approximate optimization by estimating whole-network statistics through the use of several slow, nested dynamical variables. These dynamics provide the rule with both anti-Hebbian and Hebbian components, thus allowing for decorrelating and correlating learning regimes that can occur when either is favorable for optimality. We demonstrate the performance of the synthesized rule in comparison to classical BCM dynamics and use the networks to conduct history-dependent tasks that highlight the advantages of recurrence. Finally, we show the consistency of the resultant learned networks with notions of criticality, including balanced ratios of excitation and inhibition.

EEG dynamical correlates of focal and diffuse causes of coma.

BACKGROUND: Rapidly determining the causes of a depressed level of consciousness (DLOC) including coma is a common clinical challenge. Quantitative analysis of the electroencephalogram (EEG) has the potential to improve DLOC assessment by providing readily deployable, temporally detailed characterization of brain activity in such patients. While used commonly for seizure detection, EEG-based assessment of DLOC etiology is less well-established. As a first step towards etiological diagnosis, we sought to distinguish focal and diffuse causes of DLOC through assessment of temporal dynamics within EEG signals. METHODS: We retrospectively analyzed EEG recordings from 40 patients with DLOC with consensus focal or diffuse culprit pathology. For each recording, we performed a suite of time-series analyses, then used a statistical framework to identify which analyses (features) could be used to distinguish between focal and diffuse cases. RESULTS: Using cross-validation approaches, we identified several spectral and non-spectral EEG features that were significantly different between DLOC patients with focal vs. diffuse etiologies, enabling EEG-based classification with an accuracy of 76%. CONCLUSIONS: Our findings suggest that DLOC due to focal vs. diffuse injuries differ along several electrophysiological parameters. These results may form the basis of future classification strategies for DLOC and coma that are more etiologically-specific and therefore therapeutically-relevant.

Homeostatic dynamics, hysteresis and synchronization in a low-dimensional model of burst suppression.

Burst suppression, a pattern of the electroencephalogram characterized by quasi-periodic alternation of high-voltage activity (burst) and isoelectric silence (suppression), is typically associated with states of unconsciousness, such as in deep general anesthesia and certain etiologies of coma. Recent computational models for burst suppression have attributed the slow (up to tens of seconds) time-scale of burst termination and re-initiation to cycling in supportive physiological process, such as cerebral metabolism. That is, activity-dependent substrate ('energy') depletion during bursts, followed by substrate recovery during suppression. Such a model falls into the category of a fast-slow dynamical system, commonly used to describe neuronal bursting more generally. Here, following this basic paradigm, we develop a low dimensional mean field model for burst suppression that adds several new features and capabilities to previous models. Most notably, this new model includes explicit homeostatic interactions wherein the rates of substrate recovery are tied to neuronal activity in a supply demand loop, creating a physiologically consistent, reciprocal interaction between the neural and substrate processes. We develop formal analysis of the model dynamics, showing, in particular, the capability of the model to produce burst-like activity as a consequence of neuronal downregulation only, without any direct perturbation to the substrate dynamics. Further, we use a synchronization analysis to contrast different mechanisms for spatially local versus global bursting. The analysis performed generates characterizations that are consistent with experimental observations of spatiotemporal features such as burst onset, duration, and spatial organization and, moreover, generates predictions regarding the presence of bistability and hysteresis in the underlying system. Thus, the model provides new dynamical insight into the mechanisms of burst suppression and, moreover, a tractable platform for more detailed future characterizations.

The Geometry of Plasticity-Induced Sensitization in Isoinhibitory Rate Motifs.

A well-known phenomenon in sensory perception is desensitization, wherein behavioral responses to persistent stimuli become attenuated over time. In this letter, our focus is on studying mechanisms through which desensitization may be mediated at the network level and, specifically, how sensitivity changes arise as a function of long-term plasticity. Our principal object of study is a generic isoinhibitory motif: a small excitatory-inhibitory network with recurrent inhibition. Such a motif is of interest due to its overrepresentation in laminar sensory network architectures. Here, we introduce a sensitivity analysis derived from control theory in which we characterize the fixed-energy reachable set of the motif. This set describes the regions of the phase-space that are more easily (in terms of stimulus energy) accessed, thus providing a holistic assessment of sensitivity. We specifically focus on how the geometry of this set changes due to repetitive application of a persistent stimulus. We find that for certain motif dynamics, this geometry contracts along the stimulus orientation while expanding in orthogonal directions. In other words, the motif not only desensitizes to the persistent input, but heightens its responsiveness (sensitizes) to those that are orthogonal. We develop a perturbation analysis that links this sensitization to both plasticity-induced changes in synaptic weights and the intrinsic dynamics of the network, highlighting that the effect is not purely due to weight-dependent disinhibition. Instead, this effect depends on the relative neuronal time constants and the consequent stimulus-induced drift that arises in the motif phase-space. For tightly distributed (but random) parameter ranges, sensitization is quite generic and manifests in larger recurrent E-I networks within which the motif is embedded.

Electroencephalogram signatures of loss and recovery of consciousness from propofol.

Unconsciousness is a fundamental component of general anesthesia (GA), but anesthesiologists have no reliable ways to be certain that a patient is unconscious. To develop EEG signatures that track loss and recovery of consciousness under GA, we recorded high-density EEGs in humans during gradual induction of and emergence from unconsciousness with propofol. The subjects executed an auditory task at 4-s intervals consisting of interleaved verbal and click stimuli to identify loss and recovery of consciousness. During induction, subjects lost responsiveness to the less salient clicks before losing responsiveness to the more salient verbal stimuli; during emergence they recovered responsiveness to the verbal stimuli before recovering responsiveness to the clicks. The median frequency and bandwidth of the frontal EEG power tracked the probability of response to the verbal stimuli during the transitions in consciousness. Loss of consciousness was marked simultaneously by an increase in low-frequency EEG power (<1 Hz), the loss of spatially coherent occipital alpha oscillations (8-12 Hz), and the appearance of spatially coherent frontal alpha oscillations. These dynamics reversed with recovery of consciousness. The low-frequency phase modulated alpha amplitude in two distinct patterns. During profound unconsciousness, alpha amplitudes were maximal at low-frequency peaks, whereas during the transition into and out of unconsciousness, alpha amplitudes were maximal at low-frequency nadirs. This latter phase-amplitude relationship predicted recovery of consciousness. Our results provide insights into the mechanisms of propofol-induced unconsciousness, establish EEG signatures of this brain state that track transitions in consciousness precisely, and suggest strategies for monitoring the brain activity of patients receiving GA.

A neurophysiological-metabolic model for burst suppression.

Burst suppression is an electroencepholagram (EEG) pattern in which high-voltage activity alternates with isoelectric quiescence. It is characteristic of an inactivated brain and is commonly observed at deep levels of general anesthesia, hypothermia, and in pathological conditions such as coma and early infantile encephalopathy. We propose a unifying mechanism for burst suppression that accounts for all of these conditions. By constructing a biophysical computational model, we show how the prevailing features of burst suppression may arise through the interaction between neuronal dynamics and brain metabolism. In each condition, the model suggests that a decrease in cerebral metabolic rate, coupled with the stabilizing properties of ATP-gated potassium channels, leads to the characteristic epochs of suppression. Consequently, the model makes a number of specific predictions of experimental and clinical relevance.

Thalamocortical mechanisms for the anteriorization of α rhythms during propofol-induced unconsciousness.

As humans are induced into a state of general anesthesia via propofol, the normal alpha rhythm (8-13 Hz) in the occipital cortex disappears and a frontal alpha rhythm emerges. This spatial shift in alpha activity is called anteriorization. We present a thalamocortical model that suggests mechanisms underlying anteriorization. Our model captures the neural dynamics of anteriorization when we adjust it to reflect two key actions of propofol: its potentiation of GABA and its reduction of the hyperpolarization-activated current Ih. The reduction in Ih abolishes the occipital alpha by silencing a specialized subset of thalamocortical cells, thought to generate occipital alpha at depolarized membrane potentials (>-60 mV). The increase in GABA inhibition imposes an alpha timescale on both the cortical and thalamic portions of the frontal component that are reinforced by reciprocal thalamocortical feedback. Anteriorization can thus be understood as a differential effect of anesthetic drugs on thalamic nuclei with disparate spatial projections, i.e.: (1) they disrupt the normal, depolarized alpha in posterior-projecting thalamic nuclei while (2) they engage a new, hyperpolarized alpha in frontothalamic nuclei. Our model generalizes to other anesthetics that include GABA as a target, since the molecular targets of many such anesthetics alter the model dynamics in a manner similar to that of propofol.

Local cortical dynamics of burst suppression in the anaesthetized brain.

Burst suppression is an electroencephalogram pattern that consists of a quasi-periodic alternation between isoelectric 'suppressions' lasting seconds or minutes, and high-voltage 'bursts'. It is characteristic of a profoundly inactivated brain, occurring in conditions including hypothermia, deep general anaesthesia, infant encephalopathy and coma. It is also used in neurology as an electrophysiological endpoint in pharmacologically induced coma for brain protection after traumatic injury and during status epilepticus. Classically, burst suppression has been regarded as a 'global' state with synchronous activity throughout cortex. This assumption has influenced the clinical use of burst suppression as a way to broadly reduce neural activity. However, the extent of spatial homogeneity has not been fully explored due to the challenges in recording from multiple cortical sites simultaneously. The neurophysiological dynamics of large-scale cortical circuits during burst suppression are therefore not well understood. To address this question, we recorded intracranial electrocorticograms from patients who entered burst suppression while receiving propofol general anaesthesia. The electrodes were broadly distributed across cortex, enabling us to examine both the dynamics of burst suppression within local cortical regions and larger-scale network interactions. We found that in contrast to previous characterizations, bursts could be substantially asynchronous across the cortex. Furthermore, the state of burst suppression itself could occur in a limited cortical region while other areas exhibited ongoing continuous activity. In addition, we found a complex temporal structure within bursts, which recapitulated the spectral dynamics of the state preceding burst suppression, and evolved throughout the course of a single burst. Our observations imply that local cortical dynamics are not homogeneous, even during significant brain inactivation. Instead, cortical and, implicitly, subcortical circuits express seemingly different sensitivities to high doses of anaesthetics that suggest a hierarchy governing how the brain enters burst suppression, and emphasize the role of local dynamics in what has previously been regarded as a global state. These findings suggest a conceptual shift in how neurologists could assess the brain function of patients undergoing burst suppression. First, analysing spatial variation in burst suppression could provide insight into the circuit dysfunction underlying a given pathology, and could improve monitoring of medically-induced coma. Second, analysing the temporal dynamics within a burst could help assess the underlying brain state. This approach could be explored as a prognostic tool for recovery from coma, and for guiding treatment of status epilepticus. Overall, these results suggest new research directions and methods that could improve patient monitoring in clinical practice.

Dynamical models reveal anatomically reliable attractor landscapes embedded in resting state brain networks.

Analyses of functional connectivity (FC) in resting-state brain networks (RSNs) have generated many insights into cognition. However, the mechanistic underpinnings of FC and RSNs are still not well-understood. It remains debated whether resting state activity is best characterized as noise-driven fluctuations around a single stable state, or instead, as a nonlinear dynamical system with nontrivial attractors embedded in the RSNs. Here, we provide evidence for the latter, by constructing whole-brain dynamical systems models from individual resting-state fMRI (rfMRI) recordings, using the Mesoscale Individualized NeuroDynamic (MINDy) platform. The MINDy models consist of hundreds of neural masses representing brain parcels, connected by fully trainable, individualized weights. We found that our models manifested a diverse taxonomy of nontrivial attractor landscapes including multiple equilibria and limit cycles. However, when projected into anatomical space, these attractors mapped onto a limited set of canonical RSNs, including the default mode network (DMN) and frontoparietal control network (FPN), which were reliable at the individual level. Further, by creating convex combinations of models, bifurcations were induced that recapitulated the full spectrum of dynamics found via fitting. These findings suggest that the resting brain traverses a diverse set of dynamics, which generates several distinct but anatomically overlapping attractor landscapes. Treating rfMRI as a unimodal stationary process (i.e., conventional FC) may miss critical attractor properties and structure within the resting brain. Instead, these may be better captured through neural dynamical modeling and analytic approaches. The results provide new insights into the generative mechanisms and intrinsic spatiotemporal organization of brain networks.