Across cultures and development, humans can extract proportional quantities from nonsymbolic, visual displays. Two accounts have been proposed for the contributions of these nonsymbolic skills to fraction understanding. The shared-magnitude account suggests that a cognitive system responsible for these nonsymbolic skills is foundational for symbolic fraction learning, predicting that neural patterns of fraction skills resemble magnitude-driven neural responses to continuous depictions, like unsegmented bars, that emphasize proportional magnitude. However, other depictions of proportions, like discretized, segmented bars, convey both proportional and whole-number information. The shared-interference account proposes that the interference from whole-number information in discretized proportions and symbolic fractions drives the link between these skills and predicts that neural responses to fraction magnitudes would strongly relate to overall activity to discretized depictions, particularly those with misleading whole-number information. We leveraged neural representational similarity (NRS) analyses to test these predictions. Nineteen young adults (mean age = 23.26 years) compared proportions depicted in continuous, discretized, and symbolic formats. Univariate results indicated robust neural distance effects in frontal, parietal, and occipital regions across formats. Region of interest (ROI) analyses of the intraparietal sulcus (IPS) revealed neural rational distance effects for continuous proportions, whereas for discretized and symbolic fractions, only comparisons with whole-number interference exhibited a rational magnitude signal. Critically, NRS demonstrated that the IPS fraction magnitude-related activity showed greater similarity with discretized brain responses than continuous ones, but only in the context of misleading whole-number information. Together, these findings support and refine the two accounts and suggest that symbolic fraction proficiency, as well as adept discretized proportional reasoning, involves accessing proportional magnitude code despite the presence of misleading whole-number information.

Relative to fractions, decimal numbers are thought to be easier for students to learn because they employ the same base-10 system as whole numbers. However, unlike whole numbers, larger decimals can have fewer digits, leading to worse performance when comparing Inconsistent decimal pairs, like 0.8 vs 0.26, than Consistent pairs like 0.86 vs 0.2. Students may be applying the whole number rule: “more digits = larger number” or they could be ignoring the decimal points and comparing 8 vs 26. This study used neuroimaging and our specially designed stimulus set to distinguish between these possibilities. We focused on the intraparietal sulcus (IPS), implicated in numerical magnitude processing, and the anterior cingulate cortex (ACC) and insula, implicated in inhibitory control. We found no neural differences between Consistent and Inconsistent comparisons, suggesting that the number of digits does not drive brain responses in skilled adults (n=21). Instead, for Consistent comparisons, we found that the IPS was sensitive to the actual distance between the decimals, while the ACC showed this pattern for Inconsistent comparisons. Crucially, we also examined the effect of the distance between the decimal pairs when ignoring the decimal point. Here, we found sensitivity to this distance among Inconsistent comparisons in the IPS and insula, suggesting that whole number referents are automatically processed during decimal comparison and require engagement of cognitive control regions to counteract. More broadly, our results underscore the unique challenges of decimal notation, revealing the need for educational practices that emphasize differences to whole numbers rather than highlighting similarities.

During cognitive task learning, neural representations must be rapidly constructed for novel task performance, then optimized for robust practiced task performance. How the geometry of neural representations changes to enable this transition from novel to practiced performance remains unknown. We hypothesized that practice involves a shift from compositional representations (task-general activity patterns that can be flexibly reused across tasks) to conjunctive representations (task-specific activity patterns specialized for the current task). Functional MRI during learning of multiple complex tasks substantiated this dynamic shift from compositional to conjunctive representations, which was associated with reduced cross-task interference (via pattern separation) and behavioral improvement. Further, we found that conjunctions originated in subcortex (hippocampus and cerebellum) and slowly spread to cortex, extending multiple memory systems theories to encompass cognitive task learning. The strengthening of conjunctive representations hence serves as a computational signature of learning, reflecting cortical-subcortical dynamics that optimize task representations in the human brain.

Animal and computational work indicate that phase resetting of theta oscillations (4-12 Hz) in the parahippocampal gyrus (PHG) by salient events (e.g., reward, landmarks) facilitates the encoding of goal-oriented information during navigation. Although well studied in animals, this mechanism has not been empirically substantiated in humans. In the present article, we present data from two studies (Study 1: asynchronous electroencephalography (EEG)-magnetoencephalography (MEG) | Study 2: simultaneous EEG-fMRI) to investigate theta phase resetting and its relationship with PHG blood oxygenation level dependent (BOLD) activation in healthy adults (aged 18-34 years old) navigating a virtual T-maze to find rewards. In the first experiment, both EEG and MEG data revealed a burst of theta power over right-posterior scalp locations following feedback onset (termed right-posterior theta, RPT), and RPT power and measures of phase resetting were sensitive to the subject's spatial trajectory. In Experiment 2, we used probabilistic tractography data from the human connectome project to segment the anterior and posterior PHG based on differential connectivity profiles to other brain regions. This analysis resulted in a PHG subdivision consisting of four distinct anterior and two posterior PHG clusters. Next, a series of linear mixed effects models based on simultaneous EEG-fMRI data revealed that single-trial RPT peak power significantly predicted single-trial hemodynamic responses in two clusters within the posterior PHG and one in the anterior PHG. This coupling between RPT power and PHG BOLD was exclusive to trials performed during maze navigation, and not during a similar task devoid of the spatial context of the maze. These findings highlight a role of PHG theta phase resetting for the purpose of encoding salient information during goal-directed spatial navigation. Taken together, RPT during virtual navigation integrates experimental, computational, and theoretical research of PHG function in animals with human cognitive electrophysiology studies and clinical research on memory-related disorders such as Alzheimer's disease.

Functional connectivity (FC) has been invaluable for understanding the brain's communication network, with strong potential for enhanced FC approaches to yield additional insights. Unlike with the fMRI field-standard method of pairwise correlation, theory suggests that partial correlation can estimate FC without confounded and indirect connections. However, partial correlation FC can also display low repeat reliability, impairing the accuracy of individual estimates. We hypothesized that reliability would be increased by adding regularization, which can reduce overfitting to noise in regression-based approaches like partial correlation. We therefore tested several regularized alternatives - graphical lasso, graphical ridge, and principal component regression - against unregularized partial and pairwise correlation, applying them to empirical resting-state fMRI and simulated data. As hypothesized, regularization vastly improved reliability, quantified using between-session similarity and intraclass correlation. This enhanced reliability then granted substantially more accurate individual FC estimates when validated against structural connectivity (empirical data) and ground truth networks (simulations). Graphical lasso showed especially high accuracy among regularized approaches, seemingly by maintaining more valid underlying network structures. We additionally found graphical lasso to be robust to noise levels, data quantity, and subject motion - common fMRI error sources. Lastly, we demonstrated that resting-state graphical lasso FC can effectively predict fMRI task activations and individual differences in behavior, further establishing its reliability, external validity, and ability to characterize task-related functionality. We recommend graphical lasso or similar regularized methods for calculating FC, as they can yield more valid estimates of unconfounded connectivity than field-standard pairwise correlation, while overcoming the poor reliability of unregularized partial correlation.

Rapidly learning new tasks, such as using new technology or playing a new game, is ubiquitous in our daily lives. Previous studies suggest that our brain relies on different networks for rapid task learning versus retrieving known tasks from memory, and behavioral studies have shown that novel versus practiced tasks may rely on different task configuration processes. Here, we investigated whether explicitly informing about the novelty of an incoming task would help participants prepare for different task configuration processes, such as pre-adjusting working memory gating functions. We hypothesized that if different task configuration processes can be prepared for, a pre-cue informing about the novelty of the upcoming task should lead to better task performance. Across four experiments, participants were first trained on a subset of tasks, followed by a test session in which pre-cues were provided in some blocks but not others. After comparing task performance between cued and uncued blocks, our results provided no evidence supporting the benefit of cueing for both practiced and novel tasks, suggesting that people cannot prepare for different task configuration processes in the absence of concrete task information.

A central goal of neuroscience is to understand how function-relevant brain activations are generated. Here we test the hypothesis that function-relevant brain activations are generated primarily by distributed network flows. We focused on visual processing in human cortex, given the long-standing literature supporting the functional relevance of brain activations in visual cortex regions exhibiting visual category selectivity. We began by using fMRI data from N = 352 human participants to identify category-specific responses in visual cortex for images of faces, places, body parts, and tools. We then systematically tested the hypothesis that distributed network flows can generate these localized visual category selective responses. This was accomplished using a recently developed approach for simulating - in a highly empirically constrained manner - the generation of task-evoked brain activations by modeling activity flowing over intrinsic brain connections. We next tested refinements to our hypothesis, focusing on how stimulus-driven network interactions initialized in V1 generate downstream visual category selectivity. We found evidence that network flows directly from V1 were sufficient for generating visual category selectivity, but that additional, globally distributed (whole-cortex) network flows increased category selectivity further. Using null network architectures we also found that each region's unique intrinsic "connectivity fingerprint" was key to the generation of category selectivity. These results generalized across regions associated with all four visual categories tested (bodies, faces, places, and tools), and provide evidence that the human brain's intrinsic network organization plays a prominent role in the generation of functionally relevant, localized responses.

Acute psychosocial stress affects learning, memory, and attention, but the evidence for the influence of stress on the neural processes supporting cognitive control remains mixed. We investigated how acute psychosocial stress influences performance and neural processing during the Go/NoGo task-an established cognitive control task. The experimental group underwent the Trier Social Stress Test (TSST) acute stress induction, whereas the control group completed personality questionnaires. Then, participants completed a functional magnetic resonance imaging (fMRI) Go/NoGo task, with self-report, blood pressure and salivary cortisol measurements of induced stress taken intermittently throughout the experimental session. The TSST was successful in eliciting a stress response, as indicated by significant Stress > Control between-group differences in subjective stress ratings and systolic blood pressure. We did not identify significant differences in cortisol levels, however. The stress induction also impacted subsequent Go/NoGo task performance, with participants who underwent the TSST making fewer commission errors on trials requiring the most inhibitory control (NoGo Green) relative to the control group, suggesting increased vigilance. Univariate analysis of fMRI task-evoked brain activity revealed no differences between stress and control groups for any region. However, using multivariate pattern analysis, stress and control groups were reliably differentiated by activation patterns contrasting the most demanding NoGo trials (i.e., NoGo Green trials) versus baseline in the medial intraparietal area (mIPA, affiliated with the dorsal attention network) and subregions of the cerebellum (affiliated with the default mode network). These results align with prior reports linking the mIPA and the cerebellum to visuomotor coordination, a function central to cognitive control processes underlying goal-directed behavior. This suggests that stressor-induced hypervigilance may produce a facilitative effect on response inhibition which is represented neurally by the activation patterns of cognitive control regions.

Visual shape completion is a canonical perceptual organization process that integrates spatially distributed edge information into unified representations of objects. People with schizophrenia show difficulty in discriminating completed shapes, but the brain networks and functional connections underlying this perceptual difference remain poorly understood. Also unclear is whether brain network differences in schizophrenia occur in related illnesses or vary with illness features transdiagnostically. To address these topics, we scanned (functional magnetic resonance imaging, fMRI) people with schizophrenia, bipolar disorder, or no psychiatric illness during rest and during a task in which they discriminated configurations that formed or failed to form completed shapes (illusory and fragmented condition, respectively). Multivariate pattern differences were identified on the cortical surface using 360 predefined parcels and 12 functional networks composed of such parcels. Brain activity flow mapping was used to evaluate the likely involvement of resting-state connections for shape completion. Illusory/fragmented task activation differences ('modulations') in the dorsal attention network (DAN) could distinguish people with schizophrenia from the other groups (AUCs > .85) and could transdiagnostically predict cognitive disorganization severity. Activity flow over functional connections from the DAN could predict secondary visual network modulations in each group, except in schizophrenia. The secondary visual network was strongly and similarly modulated in each group. Task modulations were dispersed over more networks in patients compared to controls. In summary, DAN activity during visual perceptual organization is distinct in schizophrenia, symptomatically relevant, and potentially related to improper attention-related feedback into secondary visual areas.

Brain activity flow models estimate the movement of task-evoked activity over brain connections to help explain network-generated task functionality. Activity flow models have been shown to accurately generate task-evoked brain activations across a wide variety of brain regions and task conditions. However, these models have had limited explanatory power, given known issues with causal interpretations of the standard functional connectivity measures used to parameterize activity flow models. We show here that functional/effective connectivity (FC) measures grounded in causal principles facilitate mechanistic interpretation of activity flow models. We progress from simple to complex FC measures, with each adding algorithmic details reflecting causal principles. This reflects many neuroscientists' preference for reduced FC measure complexity (to minimize assumptions, minimize compute time, and fully comprehend and easily communicate methodological details), which potentially trades off with causal validity. We start with Pearson correlation (the current field standard) to remain maximally relevant to the field, estimating causal validity across a range of FC measures using simulations and empirical fMRI data. Finally, we apply causal-FC-based activity flow modeling to a dorsolateral prefrontal cortex region (DLPFC), demonstrating distributed causal network mechanisms contributing to its strong activation during a working memory task. Notably, this fully distributed model is able to account for DLPFC working memory effects traditionally thought to rely primarily on within-region (i.e., not distributed) recurrent processes. Together, these results reveal the promise of parameterizing activity flow models using causal FC methods to identify network mechanisms underlying cognitive computations in the human brain.

Visual shape completion is a canonical perceptual process that integrates spatially distributed edge information into unified representations of objects. People with schizophrenia show difficulty in discerning completed shapes but the brain networks and functional connections underlying this perceptual difference remain poorly understood. Equally unclear is whether similar neural differences might occur between schizophrenia and bipolar disorder.

Visual shape completion is a canonical perceptual process that integrates spatially distributed edge information into unified representations of objects. People with schizophrenia show difficulty in discerning completed shapes but the brain networks and functional connections underlying this perceptual difference remain poorly understood. Equally unclear is whether similar neural differences arise in bipolar disorder or vary across the schizo-bipolar spectrum. To shed light on these topics, we scanned (fMRI) people with schizophrenia (SZ; n=16), bipolar disorder (BP; n=15), or no psychiatric illness (HC; n=20) during rest and during a task in which they discriminated configurations that formed or failed to form completed shapes (illusory and fragmented condition, respectively). Illusory/fragmented task activation differences ("modulations"), resting-state functional connectivity, and multivariate pattern differences were identified on the cortical surface using 360 predefined parcels and 12 functional networks composed of such parcels. Brain activity flow mapping was used to evaluate the likely involvement of resting-state connections for shape completion (Cole et al., 2016). Repeated split-half cross-validation and permutation testing revealed distinct dorsal attention network (DAN) modulations in schizophrenia (SZ vs BP: p=.0006, AUC=.86; SZ vs HC: p=.0004, AUC=.87). Across patients, DAN modulations could predict cognitive disorganization symptom severity via cross-validated linear regression and permutation testing (r=.6, p<.01). Functional connections from the dorsal attention network could model activity in the secondary visual network in each group, except among those with schizophrenia (HC: p=.004, Hedges' g=.76; BP: p=.003, g=1.03; SZ: p=.25, g=.31). Compared to controls, patients' task modulations were more variable from subject to subject and dispersed over a larger number of networks. In summary, dorsal attention network activity may underpin abnormal visual perception in schizophrenia, perhaps as a result of improper feedback into secondary visual areas. Patients with either bipolar disorder or schizophrenia may compensate for abnormal perception by idiosyncratically recruiting regions across multiple non-visual networks.

Visual shape completion is a canonical perceptual process that integrates spatially distributed edge information into unified representations of objects. People with schizophrenia show difficulty in discriminating completed shapes but the brain networks and functional connections underlying this perceptual difference remain poorly understood. Also unclear is whether similar neural differences arise in bipolar disorder or vary across the schizo-bipolar spectrum. To address these topics, we scanned (fMRI) people with schizophrenia, bipolar disorder, or no psychiatric illness during rest and during a task in which they discriminated configurations that formed or failed to form completed shapes (illusory and fragmented condition, respectively). Illusory/fragmented task activation differences ('modulations'), resting-state functional connectivity, and multivariate pattern differences were identified on the cortical surface using 360 predefined parcels and 12 functional networks composed of such parcels. Brain activity flow mapping was used to evaluate the likely involvement of resting-state connections for shape completion. Dorsal attention network modulations distinguished people with schizophrenia (AUCs>.85) and could trans-diagnostically predict cognitive disorganization severity. Activity flow over functional connections from the dorsal attention network could predict secondary visual network modulations in each group, except among those with schizophrenia. Task modulations among patients were more heterogeneous and dispersed over a larger number of networks compared to controls. In summary, abnormal dorsal attention network activity emerges during perceptual organization in schizophrenia and may be related to improper attention-related feedback into secondary visual areas. Patients with either disorder may compensate for abnormal perception by an idiosyncratic recruitment of regions across multiple non-visual networks.

A central goal of neuroscience is to understand how function-relevant brain activations are generated. Here we test the hypothesis that function-relevant brain activations are generated primarily by distributed network flows. We focused on visual processing in human cortex, given the long-standing literature supporting the functional relevance of brain activations in visual cortex regions exhibiting visual category selectivity. We began by using fMRI data from N=352 human participants to identify category-specific responses in visual cortex for images of faces, places, body parts, and tools. We then systematically tested the hypothesis that distributed network flows can generate these localized visual category selective responses. This was accomplished using a recently developed approach for simulating – in a highly empirically constrained manner – the generation of task-evoked brain activations by modeling activity flowing over intrinsic brain connections. We next tested refinements to our hypothesis, focusing on how stimulus-driven network interactions initialized in V1 generate downstream visual category selectivity. We found evidence that network flows directly from V1 were sufficient for generating visual category selectivity, but that additional, globally distributed (whole-cortex) network flows increased category selectivity further. Using null network architectures we also found that each region’s unique intrinsic “connectivity fingerprint” was key to the generation of category selectivity. These results generalized across regions associated with all four visual categories tested (bodies, faces, places, and tools), and provide evidence that the human brain’s intrinsic network organization plays a prominent role in the generation of functionally relevant, localized responses. Author Summary A fundamental question in neuroscience has persisted for over a century: to what extent do distributed processes drive brain function? The existence of category-selective regions within visual cortex provides long-standing evidence supporting localized computations, wherein specialized functions (e.g., selective responsiveness to face images) are thought to be primarily generated by within-region processes. This account was recently updated to include category selectivity dispersed across visual cortex, in the absence of category-selective regions. Here we provide groundwork evidence demonstrating that locally-exhibited visual-category-selective responses can be accurately generated via distributed activity flowing over globally connected systems. These processes were simulated via empirically-based computational models initialized by stimulus-evoked activity patterns and empirical connectivity matching each category-selective region’s unique intrinsic functional connectivity fingerprint. Results demonstrate that activity flowing over the human brain’s distributed network architecture can account for the generation of category selectivity in visual cortex regions.

How cognitive task behavior is generated by brain network interactions is a central question in neuroscience. Answering this question calls for the development of novel analysis tools that can firstly capture neural signatures of task information with high spatial and temporal precision (the "where and when") and then allow for empirical testing of alternative network models of brain function that link information to behavior (the "how"). We outline a novel network modeling approach suited to this purpose that is applied to noninvasive functional neuroimaging data in humans. We first dynamically decoded the spatiotemporal signatures of task information in the human brain by combining MRI-individualized source electroencephalography (EEG) with multivariate pattern analysis (MVPA). A newly developed network modeling approach-dynamic activity flow modeling-then simulated the flow of task-evoked activity over more causally interpretable (relative to standard functional connectivity [FC] approaches) resting-state functional connections (dynamic, lagged, direct, and directional). We demonstrate the utility of this modeling approach by applying it to elucidate network processes underlying sensory-motor information flow in the brain, revealing accurate predictions of empirical response information dynamics underlying behavior. Extending the model toward simulating network lesions suggested a role for the cognitive control networks (CCNs) as primary drivers of response information flow, transitioning from early dorsal attention network-dominated sensory-to-response transformation to later collaborative CCN engagement during response selection. These results demonstrate the utility of the dynamic activity flow modeling approach in identifying the generative network processes underlying neurocognitive phenomena.