Russell A. Poldrack - US grants

With National Science Foundation support, Dr. Poldrack will conduct a year long investigation of a new approach to enhancing brain plasticity and increasing the speed of learning in adult humans. It has long been known that the brain changes extensively early in life, and that these changes are dependent upon particular experiences in the child's environment. However, more recent research has discovered that the brain continues to change throughout adulthood in response to experience. This ability to change is called plasticity, and is thought to underlie many forms of learning. Dr. Poldrack's project will explore an approach based upon results from studies of experimental animals, which have shown that plasticity in the cerebral cortex can be greatly enhanced by increasing the levels of the neurotransmitter acetylcholine. New drugs, known as cholinesterase inhibitors, that safely and effectively increase acetylcholine levels in humans have recently been developed and FDA-approved. The specific drug that Dr. Poldrack will use is galanthamine hydrobromide (tradename Reminyl). The effect of the drug on cortical plasticity will be assessed using both visual behavioral testing and functional magnetic resonance imaging, which is a non-invasive method for measuring the brain activity that occurs as a person performs a cognitive or perceptual task. The behavioral measure will be the rate at which the subjects learn to more accurately perform a simple visual perceptual learning task: learning to discriminate the orientation of a grating. The hypothesis to be tested is that learning of the visual task that takes place under the influence of the drug will proceed more quickly than learning that is paired with a placebo. Functional magnetic resonance imaging will be used to assess the effect of the drug on cortical plasticity, by comparing the pre-training versus post-training brain activation changes that are caused by learning the visual task while on the drug against those caused by learning the task on placebo.

If this new method of enhancing plasticity should turn out to be successful, it will provide fundamentally important and novel knowledge about the nature of plasticity in the adult human brain, and could also lead to a wide range of potential clinical and practical applications. Understanding how brain plasticity works in adult humans is of critical importance, because recent research suggests that plasticity can be capitalized upon in order to remediate neurological problems, such as movement disorders resulting from stroke or from repetitive strain injury, and reading and language disorders.

DESCRIPTION (provided by applicant) Recent studies have indicated that training-induced cortical plasticity may offer promising opportunities for the remediation of several different neurological conditions, such as stroke-induced movement disorders, dystonia, and dyslexia. However, extremely intensive training has been required, and has yielded results that, although promising, fall far short of a full cure. For this reason, there has been much interest in attempts to enhance cortical plasticity, including the use of noradrenergic drugs such as amphetamine in stroke-rehabilitation, and disinhibition of motor cortex in healthy subjects using ischemic nerve block. The study proposed here will investigate a novel approach, exploiting a newly available opportunity to apply plasticity-enhancing results from the animal literature to studies in humans. Experiments in animals have shown that cortical plasticity can be greatly enhanced by increasing the levels of the neurotransmitter acetylcholine (ACh). New drugs, known as cholinesterase inhibitors, that safely and effectively increase ACh levels in humans have recently been developed and FDA-approved. The specific drug that we propose to use is galantharnine hydrobromide (trade name Reminyl). The effect of the drug on cortical plasticity will be assessed using both visual psychophysics and fMR1. The psychophysical measure will be the rate at which the subjects learn to more accurately perform a simple visual perceptual learning task: learning to discriminate the orientation of a grating. The hypothesis to be tested is that learning of the visual task that takes place under the influence of the drug will proceed more quickly than learning that is paired with a placebo. Functional MRI will be used to assess the effect of the drug on cortical plasticity, by comparing the pre-training versus post-training brain activation changes that are caused by learning the visual task while on the drug against those caused by learning the task on placebo. If the novel method of enhancing plasticity that is proposed here should turn out to be successful, then there could be a wide range of potential clinical and practical applications.

Cognitive Neuroscience of Category Learning

Abstract

With National Science Foundation support, Drs. Gluck and Poldrack and colleagues will conduct a three-year investigation to test two hypotheses about the roles of the basal ganglia and medial temporal lobe, and their interaction, in human category learning. They will test these hypotheses using a combination of functional magnetic resonance imaging (fMRI) and studies of patients with damage to either the basal ganglia (due to Parkinson's disease) or to the medial temporal lobe (due to anoxia or other causes). The first hypothesis to be tested is that the basal ganglia are particularly important for learning based on trial-by-trial feedback, whereas the medial temporal lobe is more important for observational learning in the absence of feedback. The second hypothesis is that the engagement of these two regions during category learning is modulated by the structure of the category that is being learned. For example, some categories are largely determined by single features; for example, most animals with a beak are classified as birds. Other kinds of categories require integration of information across multiple features. We will test the hypothesis that the medial temporal lobe is crucial for learning categories based on combinations of features, whereas the basal ganglia are important for learning categories based on single features.
The topic studied in this project is categorization, oneof the most important acts of human cognition. Categorization is the recognition that various individual objects can be classified into larger groups that resemble each other in some way. Research in cognitive psychology has provided a set of sophisticated models for how humans learn new categories. However, little is currently known about how these operations are achieved in the brain. Most of our current knowledge comes from studies of patients with brain disorders. In particular, patients with Parkinson's and Huntington's diseases have trouble learning some kinds of new categories, though they are able to learn other kinds of categories. These diseases affect a set of deep brain structures known as the basal ganglia, and their impairment on some category learning tasks suggests that the basal ganglia may be critical for category learning. However, the exact role of the basal ganglia is unknown. Whereas patients with Parkinson's and Huntington's diseases are severely impaired at learning some new categories, patients with amnesia following damage to the medial temporal lobe (including the hippocampus) are only subtly impaired at category learning. Initial neuroimaging studies have shown that this region is deactivated when normal individuals are learning new categories, suggesting that it is not involved. Furthermore, imaging studies have shown that that activity in the medial temporal lobe and basal ganglia during category learning is negatively related: Individuals with more activity in one region tend to have less activity in the other. These findings suggest that these two regions may interact during learning, but the nature of this interaction is unclear at present.
These studies have important implications, both for the basic understanding of category learning and for the understanding of the brain systems involved in such common disorders as Parkinson's and Alzheimer's diseases. Most importantly, the results of these studies will provide stronger constraints on theories of human category learning that are currently possible using behavioral measures alone.

This grant, funded under the Human and Social Dynamics (HSD) Competition for FY 2004, provides funding for a set of studies that use functional magnetic resonance imaging (fMRI) to examine the brain systems involved in decision making. Many decisions must be made without advance knowledge of their consequences and under some degree of physical or financial risk. Whereas the cognitive processes involved in decision making under risk are well understood, the brain systems that underlie these processes have only begun to be explored. This project will examine how brain pathways for the processing of reward and punishment are involved in decision making, by imaging brain activity while individuals make decisions regarding risky gambles (for example, the choice between a sure receipt $20 or a 50/50 chance of either $50 or $0). Factors to be examined include the relation of risky and riskless decisions, the effects of extreme probabilities on risky decision making, and the dynamics of decision making in a context of changing risk. Measures of skin conductance will be obtained during imaging, so that physiological arousal can be quantified and related to behavior and brain activity. Analyses will focus on whether activity in particular brain regions related to the experience of reward or punishment will be predictive of the individual's decisions. Each subject's behavior will also be characterized using formal decision theory models, and the parameters describing this behavior will be related to brain activity.

If successful, this research will provide novel insights into how the brain makes decisions. In particular, the results will help to relate quantitative models of behavioral decision making with their neurobiological underpinnings. Knowledge about the neural basis of decision making will encourage convergence across models at the neural and cognitive levels, and will provide neuroscientists with new ways to characterize the relation between brain activity and behavior.

DESCRIPTION (provided by applicant): The proposed research aims to develop a framework for integrating knowledge about brain, mind, and behavior across multiple disciplines, which we refer to as a }cognitive atlas. This framework extends the existing knowledge and software architecture for neuroanatomic atlas development that has been established and is now under major development in the UCLA Center for Computational Biology (CCB). A comprehensive approach to understanding mental illness requires translation between syndrome, cognition, behavior, and biology, but increasing specialization within each of these domains hinders interdisciplinary communication, insights and discoveries. The proposed project aims to bridge these gaps by developing a system for the representation of knowledge about cognitive tasks, cognitive processes, and brain structure. We propose to harness new technologies for collaborative knowledge-building along with automated methods of literature mining, to develop a flexible web-based system that will support the consolidation of distributed knowledge about cognitive phenotypes and allow creation of multi-level interdisciplinary links. This system will provide the foundation for involvement of experts in the formalization of their domain knowledge. The resulting knowledge base will provide an }atlas} that allows the mapping of cognitive phenotypes onto biomedical knowledge at manifold other levels, from genes to complex syndromes. PUBLIC HEALTH RELEVANCE: The proposed research aims to improve public health by enabling interdisciplinary research on mental disease to result in discovery of treatments and preventions that would not otherwise be possible with current approaches. It will extend the work of the National Centers for Biological Computing to develop a knowledge base that relates psychological functions to brain systems. This knowledge base spanning syndromes, cognitive processes, neural systems, and molecular genetics will additionally serve as an evolving and up-to-date educational resource for scientists, practitioners, and patients. The system developed in this project will also enhance biomedical knowledge and discovery more broadly by providing new tools for collaborative knowledge building.

DESCRIPTION (provided by applicant): This proposal requests funds to complete construction of a new home for the Imaging Research Center at the University of Texas at Austin. This center is currently located several miles from the main campus, which limits its effectiveness and availability. The proposed project would move the Imaging Research Center to a space within the newly constructed Norman Hackerman Building, located at the center of the UT Austin campus. The new location would allow the installation of additional imaging resources, including several imaging systems for small animals, as well as incorporating it more centrally into the biomedical research community. The new center will be directly connected to the building's vivarium, providing an optimal setup for animal imaging. In addition, the project would provide space for a newly purchased 3 Tesla magnetic resonance imaging system, which would support a broad range of imaging projects in humans and other species. The Norman Hackerman Building, which is replacing the Experimental Science Building of 1950, will provide UT Austin with an advanced facility for research and education in the Life Sciences. The 293,768 gross square foot building will offer modern, technology-enabled seminar rooms and undergraduate teaching laboratories, as well as office and research laboratory space for cutting-edge academic programs, including the Institute for Neuroscience, the Center for Learning and Memory, the Section of Neurobiology, and Synthetic Organic Chemistry in addition to the Imaging Research Center. The building is located within two block of nearly every biomedical research program on the UT campus. Its design incorporates the state of the art in sustainable building practices and technologies. Imaging methods have become central to nearly every aspect of biomedical research, and the availability of new imaging techniques at UT Austin has the potential to greatly advance research on campus. UT Austin has particular strength in neuroscience and cancer research, both of which would greatly benefit from the proposed improvements.

Functional magnetic resonance imaging (fMRI) has become the most common tool for cognitive neuroscience, because it provides a safe, non-invasive, and powerful means to image human brain function. Based on recent rates of publication, there are currently more than 2000 fMRI studies being performed every year worldwide. The aggregation of data across multiple studies can provide the ability to answer questions that cannot be answered based on a single study. For example, using datasets from multiple domains one can start to investigate to what degree a region is selectively engaged in relation to a particular mental process, as opposed to being generally engaged across a broad range of tasks and processes. In addition, it provides the ability to integrate across specific tasks to obtain stronger empirical generalizations about mind-brain relationships, and to better understand the nature of individual variability across different measures. Recent work in neuroimaging analysis has focused on the application of methods such as machine learning techniques to understand the coding of information at the macroscopic level, and network analysis techniques to understand the interactions inherent in large-scale neural systems. The availability of a large testbed of high-quality fMRI data from published studies would also provide an important resource for the development of these and other new analytic techniques for fMRI data. However, sharing of raw fMRI data is challenging due to the large size of the datasets and the complexity of the associated metadata, and there is currently no infrastructure for the open sharing of new fMRI datasets.

This project, OpenfMRI, will provide a new infrastructure for the broad dissemination of raw data within cognitive neuroscience, addressing a critical need by providing an open data sharing resource for neuroimaging. The initial project is already online at http://www.openfmri.org with a limited number of datasets. The full project will greatly expand this repository by providing access to a large number of fMRI datasets from several prominent neuroimaging labs, spanning across a broad range of cognitive domains. Utilizing the substantial computational resources of the Texas Advanced Computing Center, the project will also perform standard fMRI analyses on all data in the repository using a common analysis pipeline, thus providing directly comparable analysis results for all of the studies in the database. The OpenfMRI project will support the development of infrastructural elements to make sharing of data by additional investigators more straightforward.

The repository of data that will be created by the OpenfMRI project will also serve as an important resource for teaching by providing students with the ability to replicate the analyses from published studies using the same data. By providing any researcher in the world with the ability to acquire large fMRI datasets, it will also provide all researchers with the ability to work with the same state-of-the-art datasets, regardless of institution. By creating the infrastructure for open sharing of research data, the project will also enhance the impact of other NSF-funded neuroimaging research projects by providing an infrastructure that can be used to make their data available. The planned work has the potential to benefit society by improving education, health, and human productivity through an increased understanding of mental function and its relationship to brain function.

DESCRIPTION (provided by applicant): Positive changes in behavior are essential to improved health. Whereas such changes can often be instituted through intervention, long-term maintenance of these changes can be very difficult, as highlighted in this RFA. An understanding of how behaviors can be replaced and maintained is necessary in order to most effectively promote sustained beneficial changes in health-related behavior. The work proposed here will focus on understanding the brain systems that support the implementation and maintenance of behavioral change. Our work is based on the idea that there are fundamental differences between first-learned behaviors and later- learned behaviors, which result in lasting difficulties in maintenance of new behaviors in comparison to the old ones that they are meant to supplant. Although later-learned behaviors are often viewed as "replacing" first- learned ones, there is substantial behavioral evidence that first-learned behaviors are not over-written as they are extinguished, but instead are retained in a latent state. Further, it appears that maintenance of a secondary behavior requires continued suppression of the primary one in addition to acquisition of the new response. Finally, secondary behaviors appear to be substantially more context-dependent than primary ones, which results in decreased generalization to new situations. Using a combination of sophisticated neuroimaging techniques, the proposed work will characterize the brain processes that are associated with first-learned versus later-learned behaviors, and test a set of methods to improve the resilience of later-learned behaviors, with the ultimate goal of improving the long-term effectiveness of behavior change interventions. PUBLIC HEALTH RELEVANCE: Many public health problems, from obesity to heart disease to cancer to drug abuse, could be reduced by changes in behavior, but it is notoriously hard to overcome unhealthy behaviors in a lasting way. The research proposed here would investigate a previously neglected reason for this difficulty, and provide an understanding of the basic brain mechanisms that may make it difficult to overcome bad habits as well as testing several possible ways to improve the long-term effectiveness of behavioral change.

DESCRIPTION (provided by applicant): Adolescence is a risk period for the onset of drug use, with more individuals initiating drug use during these years than any other time in life. In addition, those individuals who initiate drug use earlier in adolescence are at greater risk for developing drug abuse and dependence later in life. Adolescents are known to engage disproportionately in many types of risky behavior, such as reckless driving, illicit drug use, and unsafe sex. Research has aimed to determine whether the development of neural processing could lead to the post- adolescent drop in risk-taking behavior. While several studies have demonstrated shifts in reward sensitivity across adolescence, very little research has examined the development of sensitivity to punishment. Recent research has demonstrated behaviorally that the ability to learn from punishment continues to develop throughout adolescence. It is unclear, however, if this is because children and adolescents simply find negative events less aversive than adults, or if, instead, they find punishment equally unpleasant, but are unable to incorporate this information into a learning signal, known as a prediction error signal, that is necessary to learn to approach or avoid stimuli that have previously been rewarded or punished, respectively. The proposed study aims to dissociate between these two possibilities by having groups of children, adolescents, and young adults perform a valence-modulated probabilistic learning task in an MRI scanner, where they be rewarded for correct choices during one block, and punished for incorrect choices during another. BOLD response will be measured in a contrast of negative feedback vs. positive feedback and compared across age groups to determine developmental differences in overall response to punishment, and separately prediction error response to punishment will be measured. We predict that no age differences will emerge in the negative feedback vs. positive feedback contrast, but children and adolescents will show a diminished prediction error response relative to adults.

DESCRIPTION (provided by applicant): Hispanic children are disproportionately affected by obesity and type 2 diabetes (T2D). Mounting evidence points to sugar consumption, specifically sugar-sweetened beverages (SSB), as a key modifiable factor contributing to obesity and T2D. SSB are primarily sweetened with high fructose corn syrup (HFCS), which consists of glucose and fructose in varying proportions, and the physiological responses of HFCS on metabolic health are not fully understood. Studies elucidating the mechanism of action of how HFCS impacts metabolic health, particularly in high-risk Hispanic children, are warranted. A hypothesized link between high sugar intake and metabolic disease risk involves brain reward pathways implicated in addiction. New findings with adolescents demonstrate that frequent consumption of ice cream, independent of body fat, is associated with a reduction in reward system activity, similar to the tolerance observed in drug addiction. To date, no study has examined how SSB intake impacts brain reward and addiction pathways in children, nor assessed whether this relationship differs by frequency of SSB intake. Thus, the overall goal of this study is to examine and compare how pictures of SSB and actual SSB receipt impacts brain reward pathways, food choice, subsequent food/beverage intake, and metabolic pathways in overweight Hispanic children (7-9 y), between frequent and naive drinkers. We propose a cross-sectional study of 50 Hispanic children (7-9 y), half who are frequent SSB consumers (i.e., report e 2 SSB serv/day; n=50) and half who are na¿ve drinkers (i.e., report d 2 serv/wk; n=50). Subjects will undergo two fMRI paradigms: a food choice task in which they are presented with pictures of foods/beverages that vary in their palatability/energy density and rate the value of each item, and a probabilistic reward paradigm using SSB (sweetened with HFCS) and a tasteless control. Immediately following the scan, subjects will be exposed to an ad libitum food tray, which will include a variety of healthy and unhealthy foods/beverages. Blood will be drawn at baseline (before scan), immediately after scan, and at 10-minute intervals during ad libitum food exposure to assess glucose, insulin, and gut peptides. Satiety and fullness measures will be collected at those same time points. Specific aims are: 1) to examine and compare how exposure to pictures of energy-dense foods/beverages influence choice, value, and reward pathways (striatum, amygdala, orbitofrontal cortex, insula) and examine how anticipation and receipt of SSB versus tasteless solution elicits activation of brain reward pathways between frequent and naive drinkers; and 2) to examine and compare how exposure to SSB intake in the magnet impacts subsequent acute ad libitum intake, glucose, insulin, gut peptides, and perceived satiety and hunger responses between frequent and naive drinkers. These findings will not only identify mechanisms involved in possible SSB addiction, but will also identify children most at risk for chronic sugar intake who may require more targeted interventions.

DESCRIPTION (provided by applicant): Abnormal structural and functional connectivity (interaction between brain regions) is central to the pathophysiology of psychotic illnesses like schizophrenia and psychotic bipolar disorder. Modern neuroimaging techniques and analytic strategies provide an unprecedented capacity to more fully characterize the functional and structural psychotic disconnectivity. Individuals with psychotic illness and their unaffected relatives have abnormal connectivity, suggesting that at least a portion of psychotic disconnectivity is associated with genetic predisposition for the diseases. Imaging-based connectivity endophenotypes are ideally suited to aid the functional characterization of putative risk genes, allowing us to move beyond a genotype-phenotype association to delineating mechanisms that give rise to psychotic illnesses. Recently, large-scale exome sequencing in individuals of European ancestry provided the strongest evidence to date for specific genetic variants that increase risk for psychosis. These primarily rare mutations were spread across gene networks involved in neuronal processes, including calcium channels and postsynaptic signaling. Our goals are to replicate these promising genetic findings in a different ethnic group, African-Americans, and determine whether and how these gene sets impact psychotic disconnectivity. African-Americans, an underserved population, have ~32% more highly deleterious non-synonymous rare variants in these networks than individuals of European ancestry, improving our power to detect rare variants. Our aims are to: (1) use modern MRI acquisition and analysis techniques based on the Human Connectome Project to document psychotic disconnectivity in 750 African Americans (375 with a psychotic disorder and 375 demographically matched comparison subjects). We will test hypotheses that diagnostic and dimensional indices of psychosis are associated with reduced global functional connectivity but intact global structural connectivity, combined with aberrant connectivity between specific regions or tracts; (2) conduct whole exome sequencing (WES) to test the influence of rare non-synonymous variants from genes in previously identified gene sets on psychosis risk using a network-centered analysis strategy. We will test hypotheses that the voltage-gated calcium ion channel, and the ARC-associated scaffold protein and the NMDAR postsynaptic signaling complexes influence diagnostic and dimensional indices of psychosis; and (3) apply this same network-centric test to determine if gene sets implicated in illness risk also influence functional and structural psychotic disconnectivity. Linking these genetic pathways to psychotic disconnectivity will provide mechanistic insights into the genomic influences on psychotic illness. Our collaborative application includes sites at Yale/Hartford Hospital (DC Glahn PI), Stanford (RA Poldrack PI) and Texas Biomedical Research Institute (J Blangero PI). Our results should bolster our understanding of the genetic architecture of psychotic illness and provide important clues for traversing the chasm between identified genetic networks and the behaviorally defined disorder.

General Audience Summary

This Postdoctoral Fellowship supports a research project focused on philosophical questions associated with neuroimaging technologies. The specific questions to be addressed include the following. How are new technologies for sharing, organizing, and analyzing data and theories changing the norms of evidence in neuroimaging research? How are these technologies bringing research communities, theories, data, and analysis techniques together in novel ways? What role do different data analysis techniques play in using data as evidence for claims about phenomena it was not produced to investigate? How does the use of new technologies change the way cognitive scientists interpret their data? The postdoctoral fellow will engage in participatory experience in the practice of neuroscience to obtain valuable insight into the use and structure of these technologies, which will then be used to address these philosophical questions. More specifically, the fellow will be situated in a neuroscience lab that uses and develops tools such as the Cognitive Atlas, a community driven-knowledge base, and OpenfMRI, a database of minimally processed neuroimaging data. The postdoc will contribute to ongoing research using these tools. He will gain participatory experience regarding their use and development, and while doing so engage in philosophical analysis to addresses his questions. The analysis will in turn serve to enhance the use and development of these technologies, as well as the practice of neuroscience. By situating the fellow, a philosopher of science, within the active community of cognitive neuroscientists, the project will provide members of that community with cross-disciplinary expertise. More broadly, the project will demonstrate the value of engaged philosophy of neuroscience for both advancing the practice of neuroscience, and improving the richness and quality of philosophical analyses that take neuroscience to be its subject.

Technical Summary

The proposed research project is focused on addressing philosophical issues in the epistemology of science, by approaching them from the perspective of the practices that give rise these issues. Over the last decade, neuroimaging research has involved new technologies and tools for sharing, organizing, and analyzing data and theories. This has gone hand in hand with two movements in cognitive neuroscience. The first is the growing pressure to make research practices reproducible and transparent, with the aim of improving the standards of evidence and quality of research. The second is the realization that newly developed analysis techniques that could be used to investigate hypotheses and theories previously beyond the scope of available evidence require large collections of organized and annotated data to produce meaningful results. The proposed project involves simultaneously engaging in the practice of using these technologies and studying how they are changing the evidential and theoretical landscape of cognitive neuroscience. The research fellow will use philosophical concepts to evaluate the strength of scientific inferences and to identify assumptions implicit in aspects of the technology. For example, he will explore the relations between concepts allowed by the Cognitive Atlas or its use, such as the decisions that are made during the analysis of data. This application of philosophical frameworks will also provide feedback on their descriptive and normative adequacy. In this way, this project will contribute to philosophical discussions about the epistemology of data-intensive science and neuroscience, and bring insight from those contributions to bear on the use and development of technologies (such as the Cognitive Atlas).

The Cognitive Atlas is a major online resource for testing hypotheses about the association of neural function to anatomical structures in the human brain. The goal of this project is to greatly enhance the capabilities and performance of the Atlas using state-of-the-art information technologies and national high performance computing (HPC) resources. The result will be a powerful computational tool to assist neuroscience researchers in understanding cognitive and psychological processes and making discoveries about how the brain works, in alignment with the NSF mission to promote the progress of science and to advance the national health, prosperity and welfare.

The first aim of the project is to restructure the current Cognitive Atlas database so that it will be able to interface more directly with other computing systems, and so that it can be changed in a more flexible way based on new research. The second aim is to utilize a powerful knowledge query system called Deep Dive to extract new knowledge from neuroscience publications and integrate this knowledge into the Cognitive Atlas. To address the intensive computational demands of Deep Dive, an implementation will be piloted on Wrangler, a high performance data analytics computing system hosted at the Texas Advanced Computing Center. The proposed project will enhance the computing infrastructure for neuroscience in multiple ways. By updating the system to a state-of-the-art graph database infrastructure, this work will allow much greater use of the system for automated analyses. In addition to enabling analysis of the cognitive neuroscience literature at scale for the current project, the implementation of Deep Dive on Wrangler will allow researchers in many other fields to also take advantage of this state-of-the-art application, which could have important benefits across many different areas of science and technology.

This Early-concept Grants for Exploratory Research (EAGER) award by the CISE Division of Advanced Cyberinfrastructure is jointly supported by the SBE Division of Behavioral and Cognitive Sciences and the CISE Division of Information and Intelligent Systems, with funds associated with the NSF Understanding the Brain, BRAIN Initiative activities, and for developing national research infrastructure for neuroscience. This project also aligns with NSF objectives under the National Strategic Computing Initiative.

Project Summary/Abstract The effective sharing of data requires the development and broad adoption of standards for the organization of data and metadata. Within the field of neuroimaging, there is an emerging standard for data/metadata organization, the Brain Imaging Data Structure (BIDS), which is currently implemented for MRI and is under development for PET and MEG. This standard provides the basis for organization and sharing of raw data for a broad subset of projects funded by the BRAIN Initiative. In the present project we propose to extend and expand the BIDS framework to the description and organization of data/metadata that are derived from these raw data, which we refer to generally as derivatives. This will provide the ability for researchers to share processed data, statistical models and results, and computational modeling results in a way that ensures their usability.

Project? ?Summary/Abstract The NIH BRAIN Initiative is supporting a broad portfolio of neuroscience research aimed at revolutionizing our understanding of the brain. ?The sharing of data obtained from this research is critical both to leveraging this major public investment and to ensuring the rigor and reproducibility of NIH-funded research. ?We propose to extend to the existing OpenNeuro data archive in order to provide a platform for the storage, processing, and sharing of neuroimaging data collected as part of the BRAIN Initiative. OpenNeuro extends the successful OpenfMRI project by encompassing a broader range of neuroimaging data, and by providing the ability to run data processing workflows on the data directly on the platform (using cloud computing resources) and share the results of those workflows alongside the data. The platform focuses on reproducibility through the use of versioned and containerized analysis workflows that are applied to snapshotted data releases. The proposed project would provide support for analysis and sharing of data from BRAIN Initiative projects as well as other interested projects. In addition to this infrastructure, the aims of the project are to develop a robust semi-automated curation workflow, implement new tools for federation, query, and identification, and develop support? ?for? ?advanced? ?data? ?processing? ?workflows.

Impairments in cognitive control are central to many mental health disorders (McTeague et al., 2017). In parallel, there is mounting evidence from a range of neuroimaging studies implicating impairments of network computations in disorders of mental health (Fornito et al., 2015). A crucial ?missing piece? bridging these two aspects of brain function is a relatively poor understanding of the way in which the network-level computations of the brain relate to cognitive control processes, and the precise ways in which these relationships fluctuate and unfold over weeks and months in each individual. Before we can understand fluctuations in the trajectories of mental illnesses, we need to first understand the temporal variability of healthy individuals over time. ?Recent ?dense-scanning? datasets that acquire substantially more data per subject provide a potential solution to this challenge, but these studies have lacked width (they include few subjects, e.g., 4-10) and breadth (they focus on individual tasks/states, often the ?resting state?). We will overcome these shortcoming with a dataset scanning 55 subjects each for a total 12 hours over the course of 6 months on 8 unique tasks that span multiple constructs of cognitive control (working memory, attention, set shifting, inhibition, and performance monitoring). The resultant dataset will be wide (i.e. multiple subjects per task), broad (e.g. multiple tasks per construct) and deep (e.g. multiple repetitions of each task over time). This precision neuroscience approach allows us to identify global and local changes in neural networks that are necessary both (a) in preparation for fast, effective controlled performance, and (b) to support flexible post-error and post-conflict control adjustments to improve subsequent performance. Once we have identified these behavioral and neural network signatures of cognitive control that are reproducible across task, construct, session, we will leverage this information in a novel ?targeted network attack? procedure to engineer breakdowns in the network architecture by precision challenges to the cognitive system. Tailored combinations of tasks that rely on overlapping network architectures will be combined to identify specific network features that are ripe for failure in healthy subjects, and as such, represent likely nodes for subsequent failure in disease. Together, this work will uncover novel links between cognitive control and functional brain network architecture across tasks, constructs, and sessions (Aim 1) that are essential for effective and flexible behavior (Aim 2) and are likely to fail across diverse disease states (Aim 3). Our precision neuroscience approach relates closely to the precision medicine initiative at the NIH, as our deep-scanning procedure allows us to identify subject-level network features necessary for effective cognitive control. In addition, by making the data openly accessible to other researchers, we expect these data sets will become an incomparably rich source of information for those studying the essential link between cognitive control and network-level computations.

Understanding human brain function requires knowledge of its connectivity: how one structure causally influences other components of the network. A wide range of neurological and psychiatric disorders prominently involve dysfunction of connectivity, including neurodegenerative diseases, autism, and mood disorders. Yet current methods provide only indirect measures of connectivity, and none can directly test how one brain structure causally influences another at the level of the whole brain. A unique opportunity to obtain such measures in the human brain comes from using experimental manipulation of activation through direct electrical stimulation, coupled with the whole-brain field-of-view of fMRI (es-fMRI). Our group has obtained IRB approval, and obtained strong initial data of concurrent es-fMRI in a series of 20 neurosurgical patients over the past two years. Here we intend to leverage this unique approach to the application of important open research questions in emotion, and to dissemination of protocols to a wider community of possible performance sites through this U01 mechanism. Three Aims progress through initial validation and quantification of the approach, mapping of brain networks involved in emotion processing (with a focus on the amygdala and medial prefrontal cortex), and convergent measures with ECoG and rs-fMRI. These Aims offer a mix of immediate implementation based on strong pilot data, more exploratory implementation during the grant, strong validation components, and future planning. The research focus of all Aims is on how emotion is caused by activity in brain networks. This is the topic with the strongest link to readily accessible brain structures for electrical stimulation in neurosurgical epilepsy patients (amygdala and prefrontal cortex). The work would have immediate implications for deep brain stimulation to treat diseases like depression, and long-term implications for eventually mapping out the effective functional connectome of the human brain. We will aim to provide the research community with short, feasible protocols that could be adopted by many other sites in a concerted effort to map effective connectivity in the human brain, eventually accumulating a database for understanding how individual differences in emotion, in health and disease, arise from differences in network connectivity.

To take advantage of recent and ongoing advances in intensive and large-scale computational methods, and to preserve the scientific data created by publicly funded research projects, data archives must be created as well as standards for specifying, identifying, and annotating deposited data. The value of and interest in such archives among researchers can be greatly increased by adding to them an active computational capability and framework of analysis and search tools that support further analysis as well as larger scale meta-analysis and large scale data mining. The OpenNeuro.org archive, begun as a repository for functional magnetic resonance imaging (fMRI) data, is such an archive. We propose to build a gateway to OpenNeuro for human electrophysiology data (EEG and MEG, as well as intracranial data recorded from clinical patients to plan brain surgeries or other therapies) ? herein we refer to these modalities as neuroelectromagnetic (NEM) data. The Neuroelectromagnetic Data Archive and Tools Resource (NEMAR) at the San Diego Supercomputer Center will act as a gateway to OpenNeuro for NEM data research. Such data uploaded to NEMAR at SDSC will be deposited in the OpenNeuro archive. Still- private NEM data in OpenNeuro will, on user request, be copied to the NEMAR gateway for further user processing using the XSEDE high-performance resources at SDSC in conjunction with The Neuroscience Gateway (nsgportal.org), a freely available and easy to use portal to use of high-performance computing resources for neuroscience research. Publicly available OpenNeuro NEM data will be able to be analyzed by running verified analysis applications on the OpenNeuro system. In this project we will build an application to evaluate the quality of uploaded NEM data, and another to visualize the data, for EEG and MEG at both the scalp and brain source levels, including time-domain and frequency-domain dynamics time locked to sets of experimental events learned from the BIDS- and HED-formatted data annotations. The NEMAR gateway will take a major step toward applying machine learning methods to a large store of carefully collected and stored human electrophysiologic brain data to spur new developments in basic and clinical brain research.

PROJECT SUMMARY/ABSTRACT Data sharing is essential to maximize the contributions of research subjects and the public?s investment in scientific research, but human subjects research also requires strong protection of the privacy and confidentiality of research subjects. This supplement will support an expert in neuroethics to undertake a rigorous ethical and regulatory analysis of data sharing policies, focusing in particular on the threats by artificial intelligence and machine learning techniques to reidentify neuroimaging datasets that have been thought to be deidentified. This research will lay the foundation for a sound data sharing policy for the OpenNeuro project and a regulatory framework to provide for the adequate protection of neuroimaging data while maximizing the benefits of data sharing.

Project Summary Despite the rapid advances in the neuroimaging research workflow over the last decade, the enormous variability between and within data types and specimens impedes integrated analyses. Moreover, the availability of a comprehensive portfolio of software libraries and tools has also resulted in a concerning degree of analytical variability. Generalizing the preprocessing ? that is, the intermediate step between data generation by the measurement device and the subsequent statistical modeling and analysis ? beyond fMRIPrep, we propose a framework called NiPreps (NeuroImaging Preprocessing toolS) that we envision as a workbench for the development of such pipelines. By exclusively addressing the preprocessing of the data, fMRIPrep has successfully allowed researchers to focus their effort and expertise on the portion most relevant to scientific inference (i.e., statistical and computational analyses) and reduce methodological variability. NiPreps expands fMRIPrep to operate on new imaging modalities (diffusion MRI, arterial spin labeling, positron emission tomography, and multi-echo functional MRI) and disciplines (e.g., preclinical imaging). Despite some remarkable analysis workflows that display end-to-end consolidation, integrations across applications (e.g., analyses of human and nonhuman data) remain exceptionally challenging. Hence, we will evolve fMRIPrep into NiPreps, a software framework integrating BIDS and following the BIDS-Apps specifications. First, the project will consolidate the NiPreps foundations, with the generalization of fMRIPrep's driving principles and methods across modalities and domains of application. Second, we will expand the portfolio of end-user NiPreps with dMRIPrep, ASLPrep, PETPrep, and better coverage of multi-echo fMRI by fMRIPrep. Finally, we will address the NiPreps community's consolidation to ensure the sustainability of the framework, converging the communities around each -Prep with hackathons and docusprints. In short, NIPreps will pave the way towards next-generation imaging, ultimately allowing neuroscientists to seek a unified statistical framework capable of rigorously integrating cross-application and cross-species data analysis.