Gregory S. Berns - US grants

This revised proposal for a Mentored Clinical Scientist Development Award (MCSDA) provides the applicant with an optimal scientific environment for training in the functional neuroimaging of cocaine dependence. The career development and research plans are designed to enhance the applicant's scientific skills in both positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), with the long term goals of linking changes in behavior with changes in neuronal function. In conjunction with the proposed research, the career development plan will foster the applicant's growth as an independent investigator in the cognitive neuroscience of addiction. A large body of evidence has implicated mesoaccumbal and mesolimbic circuits in cocaine dependence, yet elucidating the relationship between the function of these regions and the addiction syndrome has been hampered by the lack of knowledge about the role of these structures in human cognition. One impediment has been the absence of a functional, nonpharmacologic probe specific to the mesoaccumbal/limbic system. Targeted functional imaging paradigms, such as using working memory tasks to probe the frontal lobes, have made important advancements in the understanding of other psychiatric disorders. Based on computational models of cellular reward and information transmission, we have developed an innovative behavioral task anatomically targeted to the mesoaccumbal/limbic system. Preliminary PET data demonstrates that striatal activity normally increases in response to novel contextual information, but cocaine addicts display a decreased striatal response. The central hypothesis of this proposal is that monitoring for novelty is a normal function of the mesoaccumbal/limbic circuit, and this function becomes altered through chronic cocaine use. The preliminary findings further suggest that cocaine addicts have a different response to reinforcement, generating the secondary hypothesis that the neural substrates that code for reward will show a different response to monetary incentive. To test these hypotheses, three specific aims are identified: 1) Using PET, compare the striatal/accumbal response to novel information in a cohort of cocaine addicts and matched controls. 2) Simultaneously compare the striatal response to monetary reward in these cohorts. 3) Test the hypothesis that the striatum responds to novel information on a single-event basis by using fMRI to examine the time course of these responses. Taken together with a series of courses, seminars, and mentoring activities, this will yield important new insights into the cognitive neuroscience of human addiction and the specific roles that the mesoaccumbal and mesolimbic circuits play in cocaine dependence.

DESCRIPTION: (Adapted from the Investigator's Abstract): This project proposes a new paradigm for the study of how the dynamics of brain activity are related to behavior. An interdisciplinary bioengineering approach will be used to integrate the theory of nonlinear dynamics with both the analysis and design of functional magnetic resonance imaging (fMRI) experiments. This will lead to the development of new algorithms specifically targeted to the analysis of dynamic fMRI studies. The algorithm development will allow the implementation of novel fMRI experiments that are no longer limited to subtractive designs. In the past, the predominate approach to studying human cognition has been the subtractive experimental design. This method relies upon the establishment of discrete cognitive states, with the data analysis aimed at identifying brain regions that show significant changes in activity between the states. While successful in identifying brain regions that are differentially activated, the subtractive approach does not reveal the more complex temporal choreography that must occur to produce a specific behavior. To go beyond spatial mapping, one would like to know not only which brain regions are activated in a specific cognitive state, but how the pattern of brain activity makes the transition from one state to another. To do this, a new technique, called continuous fMRI is proposed. In contrast to the approach of designing experiments in which discrete behavioral states are maintained for blocks of time, this method maintains a single cognitive "state," but continuously varies a single parameter of the task. Pilot data from a continuously varying finger tapping task will be presented that demonstrate the feasibility of this approach. Continuous fMRI experiments generate dynamic data sets. New methods of analysis will be developed to characterize the types of dynamic behavior that occur. 1) Coupled with continuous fMRI, bifurcation analysis will identify state transitions in the brain as a single experimental parameter is continuously varied. Bifurcation theory will allow the classification of these transitions into one of four well-described forms. 2) Using finger tapping as a prototype task, bifurcation theory will be used to analyze the transition from low tapping rate to high tapping rate. 3) The technique will be extended to include the cognitive process of uncertainty detection. Using measures of information transmission, the amount of stimulus uncertainty will be varied in a reaction-time task, and bifurcation analysis will identify the types of state transitions that occur between low and high uncertainty. The integration of bifurcation theory with continuous fMRI is anticipated to have a significant impact on the way in which fMRI experiments are conducted and will yield new techniques for the study of neuropsychiatric illness.

This revised proposal for a Mentored Clinical Scientist Development Award (MCSDA) provides the applicant with an optimal scientific environment for training in the functional neuroimaging of cocaine dependence. The career development and research plans are designed to enhance the applicant's scientific skills in both positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), with the long term goals of linking changes in behavior with changes in neuronal function. In conjunction with the proposed research, the career development plan will foster the applicant's growth as an independent investigator in the cognitive neuroscience of addiction. A large body of evidence has implicated mesoaccumbal and mesolimbic circuits in cocaine dependence, yet elucidating the relationship between the function of these regions and the addiction syndrome has been hampered by the lack of knowledge about the role of these structures in human cognition. One impediment has been the absence of a functional, nonpharmacologic probe specific to the mesoaccumbal/limbic system. Targeted functional imaging paradigms, such as using working memory tasks to probe the frontal lobes, have made important advancements in the understanding of other psychiatric disorders. Based on computational models of cellular reward and information transmission, we have developed an innovative behavioral task anatomically targeted to the mesoaccumbal/limbic system. Preliminary PET data demonstrates that striatal activity normally increases in response to novel contextual information, but cocaine addicts display a decreased striatal response. The central hypothesis of this proposal is that monitoring for novelty is a normal function of the mesoaccumbal/limbic circuit, and this function becomes altered through chronic cocaine use. The preliminary findings further suggest that cocaine addicts have a different response to reinforcement, generating the secondary hypothesis that the neural substrates that code for reward will show a different response to monetary incentive. To test these hypotheses, three specific aims are identified: 1) Using PET, compare the striatal/accumbal response to novel information in a cohort of cocaine addicts and matched controls. 2) Simultaneously compare the striatal response to monetary reward in these cohorts. 3) Test the hypothesis that the striatum responds to novel information on a single-event basis by using fMRI to examine the time course of these responses. Taken together with a series of courses, seminars, and mentoring activities, this will yield important new insights into the cognitive neuroscience of human addiction and the specific roles that the mesoaccumbal and mesolimbic circuits play in cocaine dependence.

DESCRIPTION: (provided by the applicant) This project proposes the creation of a new technology that will allow the study of the biological basis of human social interaction. This technology, termed, "Hyperscan," will allow users to link magnetic resonance scanners across the internet so that functional magnetic resonance imaging (fMRI) can be performed on groups of individuals interacting with each other. The first goal of this project is to create the software platform necessary for hyperscanning. Bringing together a consortium of investigators with expertise in imaging, software development, neuroscience, and economics, we will create the software in a platform-independent manner and release it as open-source code. Additionally, we will create the architecture for locating a network of servers across the Internet to which users can connect during the performance of hyperscanning experiments. To demonstrate the utility of this approach, a single experiment drawn from the field of experimental economics will be performed. Linking together two 3 T scanners, one at Emory University and one at Princeton University, the biological substrates associated with human-human interaction will be compared to human-computer interaction.

DESCRIPTION (provided by applicant): Uncertainty has a major impact on emotional well-being. Its manifestations within psychiatric disorders are profound and range from the miscalculations of drug abuse to the ruminations of depression. Although most everyone can appreciate that uncertainty causes anxiety in normal individuals, the precise biological mechanisms by which this occurs is not known. Recent political and economic events have made it clear that we live in an unpredictable and uncertain era. The psychological effects of chronic uncertainty are powerful, and the broad aim of this proposal is to understand the neurobiological substrates of uncertainty. Using functional magnetic resonance imaging (fMRI), we propose to elucidate the brain circuitry associated with several aspects of uncertainty. We will use an objective framework for quantifying the amount of uncertainty in a particular circumstance and use several forms of Pavlovian conditioning to measure how uncertainty interacts with the brain's response. In general, we define three aspects of uncertainty: 1) uncertainty for "when"; 2) uncertainty for "what"; and 3) and higher-order (secondary) interactions. We will perform a series of simple classical conditioning experiments on healthy volunteers while they undergo fMRI scanning. We will focus on the activity in the striatal-amygdalar-hippocampal circuit, as these regions have been shown to be necessary for different forms of associative learning. More importantly, the specific roles of these structures in emotional processing are not clear, but our hypothesis is that they code the degree of uncertainty in a context-specific fashion. We will test this hypothesis with a series of 12 fMRI experiments using both pleasant and aversive stimuli. We will use variants of Pavlovian conditioning to associate pleasant oral stimuli (fruit juice) or unpleasant ones (quinine water) with different neutral cues. By varying both the predictability of temporal (when) and stimulus (what) pairings, we will be able to measure how uncertainty modulates the brain's response to this basic learning process. We hypothesize that both the amygdala and ventral striatum will be more active when uncertainty of any form is present. We will further quantify the level of this activation with autonomic measures of arousal. Finally we will investigate how these effects generalize to secondary associations. Ultimately we anticipate that this will suggest new therapies that can be targeted towards the effects of both acute and chronic uncertainty.

DESCRIPTION (provided by applicant): Adolescence represents a period of heightened vulnerability to substance use disorders. Although a variety of psychological and sociological factors contribute to this vulnerability, relatively few data exist regarding the functioning of the adolescent reward system at a neurobiological level. In this proposal, we will focus on the structures associated with the human dopaminergic system. Previous work in both animals and humans has elucidated some of the computational algorithms that the dopamine system performs. A leading candidate, the reward prediction error model, postulates that dopamine is released in response to unexpected rewards, which then modulates learning at a behavioral and synaptic level. Recent functional magnetic resonance imaging (fMRI) data from our group and others supports this model. In this application, we propose to use fMRI to assay the reactivity of the reward system in the adolescent brain. Compared to an adult, adolescent decision making capacity appears immature. Developmental models of cognition, e.g. Piaget's, suggest that immature decision making in the adolescent results from a less sophisticated stage of cognitive operation. In contrast to a staged development framework, we propose to examine the biological functioning of the reward system. Although the prediction error model of dopamine release is well supported across a range of species, the functional consequences of the developmental trajectory of the dopamine system have been relatively unexplored. To our knowledge, no data exist on the functional attributes of the adolescent reward system. If the adolescent reward system operates differently than the adult's, then this will have obvious implications for judgment and decision making, independently of cognitive operations. To accomplish this broad goal, we propose three specific aims: 1) use fMRI to measure reward system responsiviry to changes in predictability of pleasant oral stimuli, and compare the responsivity in adults and adolescents; 2) use fMRI to measure the relationship of musical preference to reward system responses in adults and adolescents to see if the adolescent reward system is more active; and 3) use fMRI to measure how peer pressure affects preference related activity in the reward system of adolescents and adults to see if the adolescent reward system is more susceptible to peer influences.

DESCRIPTION (provided by applicant): Adolescence represents a period of heightened vulnerability to substance use disorders. Although a variety of psychological and sociological factors contribute to this vulnerability, relatively few data exist regarding the functioning of the adolescent reward system at a neurobiological level. In this proposal, we will focus on the structures associated with the human dopaminergic system. Previous work in both animals and humans has elucidated some of the computational algorithms that the dopamine system performs. A leading candidate, the reward prediction error model, postulates that dopamine is released in response to unexpected rewards, which then modulates learning at a behavioral and synaptic level. Recent functional magnetic resonance imaging (fMRI) data from our group and others supports this model. In this application, we propose to use fMRI to assay the reactivity of the reward system in the adolescent brain. Compared to an adult, adolescent decision making capacity appears immature. Developmental models of cognition, e.g. Piaget's, suggest that immature decision making in the adolescent results from a less sophisticated stage of cognitive operation. In contrast to a staged development framework, we propose to examine the biological functioning of the reward system. Although the prediction error model of dopamine release is well supported across a range of species, the functional consequences of the developmental trajectory of the dopamine system have been relatively unexplored. To our knowledge, no data exist on the functional attributes of the adolescent reward system. If the adolescent reward system operates differently than the adult's, then this will have obvious implications for judgment and decision making, independently of cognitive operations. To accomplish this broad goal, we propose three specific aims: 1) use fMRI to measure reward system responsiviry to changes in predictability of pleasant oral stimuli, and compare the responsivity in adults and adolescents; 2) use fMRI to measure the relationship of musical preference to reward system responses in adults and adolescents to see if the adolescent reward system is more active; and 3) use fMRI to measure how peer pressure affects preference related activity in the reward system of adolescents and adults to see if the adolescent reward system is more susceptible to peer influences.

DESCRIPTION (provided by applicant): Current theories of drug abuse view addiction as a disease, but inevitably, the element of volitional choice will remain one of the key components of the illness. The issue boils down to this: why does an individual choose to use drugs even when they know about the negative consequences? To answer this question, we propose an experimental approach based on techniques and theories that have developed out of the field of neuroeconomics. This approach uses neuroimaging to uncover the relationships of otherwise unobservable variables in the brain that correspond to specific decisions that individuals make. Current evidence supports the theory that the orbitofrontal-striatal circuit encodes variables that relate to the economic idea of expected utility. The vast majority of these neuroeconomic studies, however, have focused on the domain of "gains." Many real-life decisions, such as the choice of whether or not to use drugs, however, also involve the potential for "loss." It is not currently known how the brain integrates decisions that involve both gains and losses. In this proposal, we will test two competing hypotheses: 1) the brain has a signed, one-dimensional system that computes the expected value of outcomes relative to some reference;or 2) the brain uses at least two different systems to compute, separately, gains and losses. To test these hypotheses, we propose 3 aims: 1) Determine the neural circuits involved in decision-making under 3 conditions: a) gains only;b) losses only;and c) gains and losses simultaneously. 2) Determine whether the circuits identified in the first aim are robust with respect to the medium of gain or loss. We will directly compare financial gains and losses with non-financial ones to test the hypothesis that a common neural "currency" is used as a measure of value for decision-making. 3) For the problem of addiction, decisions to use drugs almost always involve costs to other people as well as the individual. Using both a modified dictator game and variant of the prisoner's dilemma, we will determine whether decision-making that affects others'payoffs is also mediated by a common neural-currency hypothesis, or whether payoffs in terms of externalities on others invokes additional neural circuitry. PUBLIC HEALTH RELEVANCE: Why does an individual choose to use drugs even when they know about the negative consequences? To answer this question, we will use brain imaging to measure the brain's responses when an individual makes decisions that involve a variety of negative outcomes.

DESCRIPTION (provided by applicant): Epidemiological data point strongly to the period of late adolescence and early adulthood as the critical age during which most individuals initiate engagement in risky behaviors. Two key factors in the initiation of these behaviors are: 1) the individual's propensity to take risks (i.e. how they respond to uncertainty);and 2) the effect of social norms on individual risky decision making. In the previous period of this grant, we focused on the neurobiological underpinnings associated with an individual's response to uncertainty and the neural correlates of risky decision making. In this renewal, we propose to build on these results and address the social dimension of decision making. Using fMRI, we will determine how social messages become intertwined with the elements of uncertainty and value in the brain. The approach is grounded in neuroeconomics, which utilizes neuroimaging to uncover the relationships of hidden variables in the brain that correspond to specific decisions that individuals make. Using paradigms widely studied in experimental economics, we will test competing hypotheses about the manner in which social messages influence these types of decisions by examining the specific neural circuits that mediate the effect of social messages on levels of individual risk taking and cooperation in group settings. The overarching hypothesis is that social messages distort the expected utility of outcomes, and this will be measurable as a shift in brain activity in orbitofrontal-striatal circuits. We hypothesize this effect will be relatively insensitive to the type of message and the type of decision. To test this hypothesis, we propose 4 aims: 1) Determine the neural circuits involved in processing socially salient messages during risky decision making. 2) Determine the extent to which the source of a social message affects the specific neural circuits used to process the message. We will compare the neurobiological effects of peer-to- peer messaging, messages from authority figures and experts, and collective messages. 3) For each of the different messaging sources, we will measure its effect on 3 types of decisions that range from individual, to 2-person, to individual with consequences for others. 4) Determine the developmental trajectory of both the behavioral and neurobiological effect of social messages on risky behavior from adolescence through adulthood. PUBLIC HEALTH RELEVANCE Very little is known about how peer pressure influences risky decision making. In this project, we will use modern brain imaging techniques to see how social messages from peers and authority figures change risky decisions and which parts of the brain are responsible.