Timothy Verstynen

• Carnegie Mellon University

The modern study of perceptual learning across humans, non-human animals, and artificial agents requires large-scale datasets with flexible, customizable, and controllable features for distinguishing between categories. To support this research, we developed the Oomplet Dataset Toolkit (ODT), an open-source, publicly available toolbox capable of ge...

Although the cortico-basal ganglia-thalamic (CBGT) network is identified as a central circuit for decision-making, the dynamic interplay of multiple control pathways within this network in shaping decision trajectories remains poorly understood. Here we develop and apply a novel computational framework --- CLAW (Circuit Logic Assessed via Walks) --...

Here we introduce CBGTPy, a virtual environment for designing and testing goal-directed agents with internal dynamics that are modeled on the cortico-basal-ganglia-thalamic (CBGT) pathways in the mammalian brain. CBGTPy enables researchers to investigate the internal dynamics of the CBGT system during a variety of tasks, allowing for the formation...

Reactive inhibitory control is crucial for survival. Traditionally, this control in mammals was attributed solely to the hyperdirect pathway, with cortical control signals flowing unidirectionally from the subthalamic nucleus (STN) to basal ganglia output regions. Yet recent findings have put this model into question, suggesting that the STN is ass...

Mammalian functional architecture flexibly adapts, transitioning from integration where information is distributed across the cortex, to segregation where information is focal in densely connected communities of brain regions. This flexibility in cortical brain networks is hypothesized to be driven by control signals originating from subcortical pa...

Mammalian functional architecture flexibly adapts, transitioning from integration where information is distributed across the cortex, to segregation where information is focal in densely connected communities of brain regions. This flexibility in cortical brain networks is hypothesized to be driven by control signals originating from subcortical pa...

Reactive inhibitory control is crucial for survival. Traditionally, this control in mammals was attributed solely to the hyperdirect pathway, with cortical control signals flowing unidirectionally from the subthalamic nucleus (STN) to basal ganglia output regions. Yet recent findings have put this model into question, suggesting that the STN is ass...

For decades, the external globus pallidus (GPe) has been viewed as a passive way-station in the indirect pathway of the cortico-basal ganglia-thalamic (CBGT) circuit, sandwiched between striatal inputs and basal ganglia outputs. According to this model, one-way descending striatal signals in the indirect pathway amplify the suppression of downstrea...

It is commonplace in neuroscience to assume that if two tasks activate the same brain areas in the same way, then they are recruiting the same underlying networks. Yet computational theory has shown that the same pattern of activity can emerge from many different underlying network representations. Here we evaluated whether similarity in activation...

For decades the external globus pallidus (GPe) has been viewed as a passive way-station in the indirect pathway of the cortico-basal ganglia-thalamic (CBGT) circuit, sandwiched between striatal inputs and basal ganglia outputs. According to this model, one-way descending striatal signals in the indirect pathway amplify the suppression of downstream...

Introduction Physical activity (PA) has beneficial effects on brain health and cardiovascular disease (CVD) risk. Yet, we know little about whether PA-induced changes to physiological mediators of CVD risk influence brain health and whether benefits to brain health may also explain PA-induced improvements to CVD risk. This study combines neurobiolo...

Making adaptive choices in dynamic environments requires flexible decision policies. Previously, we showed how the evidence accumulation process that drives decisions shifts when outcome contingencies change (1). Using in silico experiments, here we show how the cortico-basal ganglia-thalamic (CBGT) circuits can feasibly implement shifts in the evi...

The ability to change your mind when the local environment changes relies critically on cortico-basal ganglia-thalamic (CBGT) circuits. In silico experiments on the CBGT pathways show how shifts in decision policy are driven by learning-induced changes in competition between action plans, both within and across action representations. We empiricall...

In situations featuring uncertainty about action-reward contingencies, mammals can flexibly adopt strategies for decision-making that are tuned in response to environmental changes. Although the cortico-basal ganglia thalamic (CBGT) network has been identified as contributing to the decision-making process, it features a complex synaptic architectu...

Repetition of specific movement biases subsequent actions towards the practiced movement, a phenomenon known as use-dependent learning (UDL). Recent experiments that impose strict constraints on planning time have revealed two sources of use-dependent biases, one arising from dynamic changes occurring during motor planning and another reflecting a...

In the basal ganglia, different dopamine subtypes have opposing dynamics at post-synaptic receptors, with the ratio of D1 to D2 receptors determining the relative sensitivity to gains and losses, respectively, during value-based learning. This effective sensitivity to reward feedback interacts with phasic dopamine levels to determine the effectiven...

In uncertain or unstable environments, sometimes the best decision is to change your mind. To shed light on this flexibility, we evaluated how the underlying decision policy adapts when the most rewarding action changes. Human participants performed a dynamic two-armed bandit task that manipulated the certainty in relative reward (conflict) and the...

During action selection, mammals exhibit a high degree of flexibility in adapting their decisions in response to environmental changes. Although the cortico-basal ganglia thalamic (CBGT) network is implicated in this adaptation, it features a synaptic architecture comprising multiple feed-forward, reciprocal, and feedback pathways, complicating eff...

There is an ongoing debate as to whether cognitive processes arise from a group of functionally specialized brain modules (modularism) or as the result of a distributed nonlinear process (dynamical systems theory). The former predicts that tasks that recruit similar brain areas should have an equivalent degree of similarity in their connectivity. T...

While stress may be a potential mechanism by which childhood threat and deprivation influence mental health, few studies have considered specific stress‐related white matter pathways, such as the stria terminalis (ST) and medial forebrain bundle (MFB). Our goal was to examine the relationships between childhood adversity and ST and MFB structural i...

Variation in cognitive ability arises from subtle differences in underlying neural architecture. Understanding and predicting individual variability in cognition from the differences in brain networks requires harnessing the unique variance captured by different neuroimaging modalities. Here we adopted a multi-level machine learning approach that c...

Humans and other mammals flexibly select actions under noisy and unstable conditions. To shed light on the mechanism driving this flexibility, we evaluated how the underlying decision policy evolves when humans change their minds about the most rewarding action. Participants performed a dynamic variant of the two-armed bandit task that manipulated...

Variation in cognitive ability arises from subtle differences in underlying neural architectural properties. Understanding and predicting individual variability in cognition from the differences in brain networks requires harnessing the unique variance captured by different neuroimaging modalities. Here we adopted a multi-level machine learning app...

Unhealthy weight gain relates, in part, to how people make decisions based on prior experience. Here we conducted post hoc analysis on an archival data set to evaluate whether individual differences in adiposity, an anthropometric construct encompassing a spectrum of body types, from lean to obese, associate with signatures of asymmetric feedback l...

Deep reinforcement learning can match or exceed human performance in stable contexts, but with minor changes to the environment artificial networks, unlike humans, often cannot adapt. Humans rely on a combination of heuristics to simplify computational load and imagination to extend experiential learning to new and more challenging environments. Mo...

Although the same decision to act can occur in multiple contexts, how these contexts differentially influence behavior is not well understood. In this paper, we investigate whether contextual framing affects individuals' behavior in spatial decision making. Although previous research suggests that individuals' judgments are sensitive to contextual...

The question of how cortico‐basal ganglia‐thalamic (CBGT) pathways use dopaminergic feedback signals to modify future decisions has challenged computational neuroscientists for decades. Reviewing the literature on computational representations of dopaminergic corticostriatal plasticity, we show how the field is converging on a normative, synapticle...

In natural environments, mammals can efficiently select actions based on noisy sensory signals and quickly adapt to unexpected outcomes to better exploit opportunities that arise in the future. Such feedback-based changes in behavior rely, in part, on long term plasticity within cortico-basal-ganglia-thalamic networks, driven by dopaminergic modula...

Value-based decision-making relies on effective communication across disparate brain networks. Given the scale of the networks involved in adaptive decision-making, variability in how they communicate should impact behavior; however, precisely how the topological pattern of structural connectivity of individual brain networks influences individual...

Value-based decision-making relies on the communication across disparate brain networks. Given the scale of the networks involved in adaptive decision-making, variability in how they communicate should impact behavior; however, precisely how the topological pattern of structural connectivity of individual brain networks influences individual differ...

As a sequence of movements is learned, serially ordered actions get bound together into sets to reduce computational complexity during planning and execution. Here, we investigated how actions become naturally bound over the course of learning and how this learning affects cortical representations of individual actions. Across 5 weeks of practice,...

Community detection algorithms have been widely used to study the organization of complex networks like the brain. These techniques provide a partition of brain regions (or nodes) into clusters (or communities), where nodes within a community are densely interconnected with one another. In their simplest application, community detection algorithms...

Distribution of edge weights in the structural and functional brain graphs. The distribution of streamline counts for all edges in the structural brain graphs at the subject level (A) and at the group level (B). Note the heavy-tailed distribution of streamline counts where hundreds of edges are made up of only a few streamlines while only a handful...

Principal components analysis of the SC and FC stability matrices extracted from a single representative subject. (A) The SC stability matrix and (B) FC stability matrix created by exchanging the community labels with their corresponding stability values. (C&D) Principal components analysis of the matrices shown in panels (A&B), with eight principa...

Synthetic graphs with diverse hierarchical structure. The adjacency matrices (Left) and their corresponding hierarchical community structure identified using the multi-scale method (Right) are presented for three different graphs (A,C,E) where nodes form hierarchical communities at three separate topological scales. In graph (A), the communities at...

Group consensus hierarchical community organization of functional graphs. The communities at several γ values are color-coded and overlaid separately for each scale. Singleton communities are removed from the visualization for clarity. (EPS)

Group consensus hierarchical community organization of structural graphs. The communities at several γ values are color-coded and overlaid separately for each scale. Singleton communities are removed from the visualization for clarity. (EPS)

Effect of the topological scale coupling parameter on the identified hierarchical community structure. (A-F) Panels show the multi-scale communities of the synthetic graph in Fig 2 of the main text, detected when using different values of the topological scale coupling parameter (τ). Here we similarly used a γ ∈ [0, 12], an inter-layer γ increment...

Spatially disconnected communities of the group-level hierarchical community structure in structural brain graphs. Only the spatially disconnected communities (i.e., multi-cluster communities) at several γ values are color-coded and overlaid separately for each scale (for details, see section 1 in the appendix). Singleton communities are removed. N...

Hierarchical community organization of a multiplex, multi-scale graph that explicitly combines both imaging modalities: Structure and function. (A) The multiplex graph representing both imaging modalities is produced by coupling the nodes with the same identity across modalities (i.e., across FC and SC graphs). The strength of the coupling between...

Similarity between the allegiance matrices of the multiplex SC-FC graph, the FC graph alone, and the SC graph alone, as a function of the topological scale for a representative subject. (A-B) Pearson correlation coefficient values for all layers and all layer shifts between the allegiance matrices of the multiplex SC-FC graph and the allegiance mat...

Multi-scale community organization of a synthetic dynamic graph. (A) We created a dynamic graph by periodically changing the community structure of the graph. In this example, the network starts by switching between two community structures, indicated by the colors red and green. However, at slower time scales the switching dynamics between the red...

Number of spatially disconnected communities of the group-level hierarchical community structure in structural brain graphs. The total number of communities and the total number of spatially disconnected communities (i.e., multi-cluster communities) at each γ value are marked by ’o’ and ’*’, respectively. The null distributions (N = 1000) of the to...

Cortico-basal-ganglia-thalamic (CBGT) networks are critical for adaptive decision-making, yet how changes to circuit-level properties impact cognitive algorithms remains unclear. Here we explore how dopaminergic plasticity at corticostriatal synapses alters competition between striatal pathways, impacting the evidence accumulation process during de...

The human brain is a complex dynamical system, and how cognition emerges from spatiotemporal patterns of regional brain activity remains an open question. As different regions dynamically interact to perform cognitive tasks, variable patterns of partial synchrony can be observed, forming chimera states. We propose that the spatial patterning of the...

Goal-directed behavior requires integrating action selection processes with learning systems that adapt control using environmental feedback. These functions are known to intersect at a common neural substrate with multiple known targets of plasticity (the cortico-basal ganglia-thalamic network), suggesting that feedback signals have a multifaceted...

In the real world, agents often have to operate in situations with incomplete information, limited sensing capabilities, and inherently stochastic environments, making individual observations incomplete and unreliable. Moreover, in many situations it is preferable to delay a decision rather than run the risk of making a bad decision. In such situat...

Cortico-basal-ganglia-thalamic (CBGT) networks are critical for adaptive decision-making, yet how changes to circuit-level properties impact cognitive algorithms remains unclear. Here we explore how dopaminergic plasticity at corticostriatal synapses alters competition between striatal pathways, impacting the evidence accumulation process during de...

Deep reinforcement learning can match and exceed human performance, but if even minor changes are introduced to the environment artificial networks often can't adapt. Humans meanwhile are quite adaptable. We hypothesize that this is partly because of how humans use heuristics, and partly because humans can imagine new and more challenging environme...

The human brain is a complex dynamical system that gives rise to cognition through spatiotemporal patterns of coherent and incoherent activity between brain regions. As different regions dynamically interact to perform cognitive tasks, variable patterns of partial synchrony can be observed, forming chimera states. We propose that the emergence of s...

Health behaviors arise from the dynamics of highly interconnected networks in the brain and variability in these networks drives individual differences in behavior. In this review, we show how many factors that predict the physical health of the body also correlate with variability of the myelinated fascicles, called white matter, that connect brai...

A comprehensive map of the structural connectome in the human brain has been a coveted resource for understanding macroscopic brain networks. Here we report an expert-vetted, population-averaged atlas of the structural connectome derived from diffusion MRI data (N = 842). This was achieved by creating a high-resolution template of diffusion pattern...

Large bundles of myelinated axons, called white matter, anatomically connect disparate brain regions together and compose the structural core of the human connectome. We recently proposed a method of measuring the local integrity along the length of each white matter fascicle, termed the local connectome. If communication efficiency is fundamentall...

The unique architecture of the human connectome is defined initially by genetics and subsequently sculpted over time with experience. Thus, similarities in predisposition and experience that lead to similarities in social, biological, and cognitive attributes should also be reflected in the local architecture of white matter fascicles. Here we empl...

As a movement sequence is learned, serially ordered actions get bound together into sets in order to reduce computational complexity during planning and execution. Here we examined how the binding of serial actions alters the cortical representations of individual movements. Across five weeks of practice, healthy human subjects learned either a com...

Recent advances in deep learning have allowed artificial agents to rival human-level performance on a wide range of complex tasks; however, the ability of these networks to learn generalizable strategies remains a pressing challenge. This critical limitation is due in part to two factors: the opaque information representation in deep neural network...

A comprehensive map of the structural connectome in the human brain has been a coveted resource for understanding macroscopic brain networks. Here we report an expert-vetted, population-averaged atlas of the structural connectome derived from diffusion MRI data (N=842). This was achieved by creating a high-resolution template of diffusion patterns...

Goal-directed behavior requires integrating action selection processes with learning systems that adapt control using environmental feedback. These functions intersect in the basal ganglia (BG), which has at least two targets of plasticity: a dopaminergic modulation of striatal pathways and cortical modulation of the subthalamic nucleus (STN). Dual...

Goal-directed behavior requires integrating action selection processes with learning systems that adapt control using environmental feedback. These functions intersect in the basal ganglia (BG), which has at least two targets of plasticity: a dopaminergic modulation of striatal pathways and cortical modulation of the subthalamic nucleus (STN). Dual...

The unique architecture of the human connectome is defined initially by genetics and subsequently sculpted over time with experience. Thus, similarities in predisposition and experience that lead to similarities in social, biological, and cognitive attributes should also be reflected in the local architecture of white matter fascicles. Here we empl...

Obesity is associated with functional and structural differences in the corticostriatal systems of the brain. These pathways are known to be critical for the acquisition of complex sensorimotor skills, such as the ability to learn a coordinated sequence of actions. Thus, individual differences in obesity should be associated with reduced efficiency...

Community detection algorithms have been widely used to study the organization of complex systems like the brain. A principal appeal of these techniques is their ability to identify a partition of brain regions (or nodes) into communities, where nodes within a community are densely interconnected. In their simplest application, community detection...

The dual-system model of sequence learning posits that during early learning there is an advantage for encoding sequences in sensory frames; however, it remains unclear whether this advantage extends to long-term consolidation. Using the serial RT task, we set out to distinguish the dynamics of learning sequential orders of visual cues from learnin...

In the last few decades, noninvasive neuroimaging has revealed macroscale brain dynamics that underlie perception, cognition, and action. Advances in noninvasive neuroimaging target two capabilities: 1) increased spatial and temporal resolution of measured neural activity; and 2) innovative methodologies to extract brain–behavior relationships from...

Post-task resting state dynamics can be viewed as a task-driven state where behavioral performance is improved through endogenous, non-explicit learning. Tasks that have intrinsic value for individuals are hypothesized to produce post-task resting state dynamics that promote learning. We measured simultaneous fMRI/EEG and DTI in Division-1 collegia...

The flexibility of behavioral control is a testament to the brain’s capacity for dynamically resolving uncertainty during goal-directed actions. This ability to select actions and learn from immediate feedback is driven by the dynamics of basal ganglia (BG) pathways. A growing body of empirical evidence conflicts with the traditional view that thes...

The flexibility of behavioral control is a testament to the brain's capacity for dynamically resolving uncertainty during goal-directed actions. This ability to select actions and learn from immediate feedback is driven by the dynamics of basal ganglia (BG) pathways. A growing body of empirical evidence conflicts with the traditional view that thes...

Here we introduce the concept of the local connectome: the degree of connectivity between adjacent voxels within a white matter fascicle defined by the density of the diffusing spins. While most human structural connectomic analyses can be summarized as finding global connectivity patterns at either end of anatomical pathways, the analysis of local...

Table of significant clusters for the no-go parametric minus go parametric contrast shown in Figure 8A. Coordinates are centers of mass for the cluster in MNI-space. N is the number of voxels in each cluster. Values in the left six columns show average condition-wise (general linear model) GLM coefficients and standard deviation across subjects is...

The architecture of corticobasal ganglia pathways allows for many routes to inhibit a planned action: the hyperdirect pathway performs fast action cancellation and the indirect pathway competitively constrains execution signals from the direct pathway. We present a novel model, principled off of basal ganglia circuitry, that differentiates control...

Projections from the substantia nigra and striatum traverse through the pallidum on the way to their targets. To date, in vivo characterization of these pathways remains elusive. Here we used high angular resolution diffusion imaging (N=138) to study the characteristics and structural subcompartments of the human pallidum. Our central result shows...

Projections from the substantia nigra and striatum traverse through the pallidum on the way to their targets. To date, in vivo characterization of these pathways remains elusive. Here we used high angular resolution diffusion imaging (N=138) to study the characteristics and structural subcompartments of the human pallidum. Our results show that the...

Modification of spatial attention via reinforcement learning (Lee & Shomstein, 2013) requires the integration of reward, attention, and executive processes. Corticostriatal pathways are an ideal neural substrate for this integration because these projections exhibit a globally parallel (Alexander, De Long, and Strick, 1985), but locally overlapping...

Being overweight or obese is associated with reduced white matter integrity throughout the brain. It is not yet clear which physiological systems mediate the association between inter-individual variation in adiposity and white matter. We tested whether composite indicators of cardiovascular, lipid, glucose, and inflammatory factors would mediate t...

A network of multiple brain regions is recruited in face perception. Our understanding of the functional properties of this network can be facilitated by explicating the structural white matter connections that exist between its functional nodes. We accomplished this using functional MRI (fMRI) in combination with fiber tractography on high angular...

The middle longitudinal fascicle (MdLF) was originally described in the monkey brain as a pathway that interconnects the superior temporal and angular gyri. Only recently have diffusion tensor imaging studies provided some evidence of its existence in humans, with a connectivity pattern similar to that in monkeys and a potential role in the languag...

Socioeconomic disadvantage confers risk for aspects of ill health that may be mediated by systemic inflammatory influences on the integrity of distributed brain networks. Following this hypothesis, we tested whether socioeconomic disadvantage related to the structural integrity of white matter tracts connecting brain regions of distributed networks...

High-definition fiber tracking (HDFT) is a novel combination of processing, reconstruction, and tractography methods that can track white matter fibers from cortex, through complex fiber crossings, to cortical and subcortical targets with subvoxel resolution. To perform neuroanatomical validation of HDFT and to investigate its neurosurgical applica...

For patients with traumatic brain injury (TBI), current clinical imaging methods generally do not provide highly detailed information about the location of axonal injury, severity of injury, or expected recovery. In a case of severe TBI, the authors applied a novel high-definition fiber tracking (HDFT) to directly visualize and quantify the degree...

The axons that project into the striatum are known to segregate according to macroscopic cortical systems; however, the within-region organization of these fibers has yet to be described in humans. We used in vivo fiber tractography, in neurologically healthy adults, to map white matter bundles that originate in different neocortical areas, navigat...

Visual attention selects behaviorally relevant information for detailed processing by resolving competition for representation among stimuli in retinotopically organized visual cortex. The signals that control this attentional biasing are thought to arise in a frontoparietal network of several brain regions, including posterior parietal cortex. Rec...

Histological studies on nonhuman primates have shown a rich topography of homotopic (i.e., going to the same regions) or heterotopic (i.e., going to different regions) callosal projections. Unfortunately, a complete within-subject mapping of commissural projections in humans has been limited due to the inability of typical imaging methods to detect...

Executing difficult actions with the left hand results in bilateral activity of motor areas along the precentral gyrus. Using TMS and fMRI, we explored the functional relationship between primary (M1) and premotor areas during unimanual actions, focusing on M1 activity in the ipsilateral hemisphere. Single-pulse TMS revealed that the amplitude of m...

Most voluntary actions rely on neural circuits that map sensory cues onto appropriate motor responses. One might expect that for everyday movements, like reaching, this mapping would remain stable over time, at least in the absence of error feedback. Here we describe a simple and novel psychophysical phenomenon in which recent experience shapes the...

We have recently shown that the statistical properties of goal directed reaching in human subjects depends on recent experience in a way that is consistent with the presence of adaptive Bayesian priors (Verstynen and Sabes, 2011). We also showed that when Hebbian (associative) learning is added to a simple line-attractor network model, the network...

The BOLD signal not only reflects changes in local neural activity, but also exhibits variability from physiological processes like cardiac rhythms and breathing. We investigated how both of these physiological sources are reflected in the pulse oximetry (PO) signal, a direct measure of blood oxygenation, and how this information can be used to acc...