Paul W. Glimcher - US grants

DESCRIPTION (Adapted From The Applicant's Abstract): The long-term goal of our research is to determine exactly what role the substantia nigra plays in the generation of normal eye movements. While previous research has focused on how neurons in the reticulata subdivision of this nucleus may influence the generation of saccades by the superior colliculus, very little effort has been focused on attempting to understand how both the reticulata and compacta subdivisions might participate in the movement control process in general. Over the past several years our laboratory has begun to approach this goal by examining the activity of reticulata neurons. While we have confirmed earlier findings that some reticulata neurons are suppressed when subjects plan and initiate movements for which they receive rewards, we have also found that reticulata neurons are unmodulated when subjects produce movements for which they are not rewarded. These data suggest that reticulata activity is more tightly coupled to sequences of movements that are reinforced, and hence adaptive for the subjects, than with movements in general. Existing studies of the compacta subdivision seem to indicate a similar functional role for that area, linking compacta rate to the adaptive value of behavioral responses. The goal of this proposal is to test the hypothesis that both pars compacta and pars reticulata carry signals related to both the adaptive significance of stimuli that guide movements and to the reinforced movements themselves, not simply signals locked to the precise stimuli the animal perceives or the precise movements the animal executes. To examine this question, and to identify precisely what role reticulata and compacta signals might play in oculomotor processes, we propose two sets of interlocking physiological experiments: one set directed at the pars reticulata and a second set directed at the pars compacta. The reticulata experiments proposed here are designed to determine: whether reticulata task-related modulations can be elicited by non-visual stimuli, whether saccade-related reticulata modulations can be associated with entire sequences of movements, whether reinforcers not associated with eye movements modulate reticulata neurons, and whether changes in the relevance of a stimulus to a behavioral task can alter the reticulata modulation associated with that event. The compacta experiments proposed here test the hypothesis that the two nigral subnuclei are both involved in the generation of adaptively valuable movements by directly comparing the responses of the compacta and reticulata neurons under identical conditions, while subjects perform eye movements for which they receive reinforcement. These experiments may ultimately tell us how this nucleus, which has been implicated in a number of human disorders, participates in the generation of movements by normal individuals.

DESCRIPTION (Adapted From The Applicant's Abstract): The long-term goal of our research is to determine exactly what role the substantia nigra plays in the generation of normal eye movements. While previous research has focused on how neurons in the reticulata subdivision of this nucleus may influence the generation of saccades by the superior colliculus, very little effort has been focused on attempting to understand how both the reticulata and compacta subdivisions might participate in the movement control process in general. Over the past several years our laboratory has begun to approach this goal by examining the activity of reticulata neurons. While we have confirmed earlier findings that some reticulata neurons are suppressed when subjects plan and initiate movements for which they receive rewards, we have also found that reticulata neurons are unmodulated when subjects produce movements for which they are not rewarded. These data suggest that reticulata activity is more tightly coupled to sequences of movements that are reinforced, and hence adaptive for the subjects, than with movements in general. Existing studies of the compacta subdivision seem to indicate a similar functional role for that area, linking compacta rate to the adaptive value of behavioral responses. The goal of this proposal is to test the hypothesis that both pars compacta and pars reticulata carry signals related to both the adaptive significance of stimuli that guide movements and to the reinforced movements themselves, not simply signals locked to the precise stimuli the animal perceives or the precise movements the animal executes. To examine this question, and to identify precisely what role reticulata and compacta signals might play in oculomotor processes, we propose two sets of interlocking physiological experiments: one set directed at the pars reticulata and a second set directed at the pars compacta. The reticulata experiments proposed here are designed to determine: whether reticulata task-related modulations can be elicited by non-visual stimuli, whether saccade-related reticulata modulations can be associated with entire sequences of movements, whether reinforcers not associated with eye movements modulate reticulata neurons, and whether changes in the relevance of a stimulus to a behavioral task can alter the reticulata modulation associated with that event. The compacta experiments proposed here test the hypothesis that the two nigral subnuclei are both involved in the generation of adaptively valuable movements by directly comparing the responses of the compacta and reticulata neurons under identical conditions, while subjects perform eye movements for which they receive reinforcement. These experiments may ultimately tell us how this nucleus, which has been implicated in a number of human disorders, participates in the generation of movements by normal individuals.

biomedical equipment; vision tests; biomedical facility; clinical biomedical equipment; bioengineering /biomedical engineering; biomedical equipment development; computer system hardware; psychophysics; computer program /software; online computer; electrodes; visual tracking; health science research support;

DESCRIPTION (provided by applicant): While accumulating evidence now supports the notion that several anatomical targets of the basal ganglia play a critical role in eye movement generation by making decisions both about what movement to make and when to make those movements, studies in the basal ganglia are beginning to suggest a different role for these nuclei. Several lines of research now suggest that the basal ganglia play an evaluative role, assessing whether recently completed movements were of value to an organism. If this is true, then movement disorders associated with basal ganglia dysfunction must be viewed as fundamentally different from movement disorders associated with damage to areas involved in movement execution. A series of experiments are proposed to test this hypothesis. First, the proposal attempts to determine whether eye movement-related neurons in the substantia nigra pars compacta signal when an unexpectedly positive outcome has occurred, when an unexpected reward has been obtained by the organism. Second, the proposal attempts to test the hypothesis that caudate neurons use this information to determine what specific movements may have led to the unexpectedly positive outcome. Three sets of experiments engage these two hypotheses: i) Do neurons of the pars compacta encode the time and value of an unexpected reward (as quantitatively predicted by theories of reinforcement learning) but do not carry information about the metrical properties of recent movements? A set of single-unit recording studies, driven by existing theoretical models, engage this question, ii) Do saccade-related neurons in the caudate nucleus carry a signal that determines which eye movements account for the unexpected rewards signaled by compacta activity? A set of single-unit recording studies driven by recent results in the caudate engage this question. iii) Does artificial activation of compacta cause the caudate to produce new links between movements and rewards, and if so what is the temporal signature of this attributional process? A combination of single-unit recording and microstimulation techniques seek to causally link compacta activation to both behavioral responses and caudate activity patterns. Together these experiments would help provide a new circuit-level overview of the function of the basal ganglia, an overview which would have a significant impact on both future basic research and clinical models of basal ganglia function in eye movement control.

DESCRIPTION (provided by applicant): While accumulating evidence now supports the notion that several anatomical targets of the basal ganglia play a critical role in eye movement generation by making decisions both about what movement to make and when to make those movements, studies in the basal ganglia are beginning to suggest a different role for these nuclei. Several lines of research now suggest that the basal ganglia play an evaluative role, assessing whether recently completed movements were of value to an organism. If this is true, then movement disorders associated with basal ganglia dysfunction must be viewed as fundamentally different from movement disorders associated with damage to areas involved in movement execution. A series of experiments are proposed to test this hypothesis. First, the proposal attempts to determine whether eye movement-related neurons in the substantia nigra pars compacta signal when an unexpectedly positive outcome has occurred, when an unexpected reward has been obtained by the organism. Second, the proposal attempts to test the hypothesis that caudate neurons use this information to determine what specific movements may have led to the unexpectedly positive outcome. Three sets of experiments engage these two hypotheses: i) Do neurons of the pars compacta encode the time and value of an unexpected reward (as quantitatively predicted by theories of reinforcement learning) but do not carry information about the metrical properties of recent movements? A set of single-unit recording studies, driven by existing theoretical models, engage this question, ii) Do saccade-related neurons in the caudate nucleus carry a signal that determines which eye movements account for the unexpected rewards signaled by compacta activity? A set of single-unit recording studies driven by recent results in the caudate engage this question. iii) Does artificial activation of compacta cause the caudate to produce new links between movements and rewards, and if so what is the temporal signature of this attributional process? A combination of single-unit recording and microstimulation techniques seek to causally link compacta activation to both behavioral responses and caudate activity patterns. Together these experiments would help provide a new circuit-level overview of the function of the basal ganglia, an overview which would have a significant impact on both future basic research and clinical models of basal ganglia function in eye movement control.

Which would you prefer, a gain of $20 in 12 months or of $22 in 13 months? Most people would wait the extra month to gain the extra $2. Ask yourself the same question 365 days later: which would you prefer, $20 today or $22 in a month? Now most people would select the immediate gain of $20;a reversal in preferences. Sitting in a clinic, most drug users state a preference for avoiding drug use later in the day. Hours later, confronting an immediately available drug, their preferences reverse. These observations reveal that humans make choices by comparing the subjective values of rewards and that these values depend not just on how large a reward is but when that reward will be received. The longer gratification is delayed;the less desirable a reward appears. Empirical data suggeststhat this decrease in desirability occurs most abruptly at short delays and more gradually as delays increase. Perhaps just as interesting is the observation that essentially all vertebrates appear to show a similar pattern. What then, is the neural mechanism that sets a subjective value on delayed rewards? Neurobiological studies of decision-making have largely avoided this issue. In this application we propose to extend existing neurobiological studies of decision-making into the domain of choice-in-time. We propose to undertake this study with a combination of behavioral, physiological and functional magnetic resonance imaging techniques and to test two hypotheses that seek seek to explain choice-in-time: a standard physiological model and the dominant two-system model of economic theory. Specifically, we propose to test the hypothesis that either of these models can account for the activity of the posterior parietal cortex and the ventral striatum under a range of conditions. Pathological influences of delays on decision making are pervasive in disease. Delays distort the decisions made by children with attention-deficit disorder, pathological gamblers, drug abusers, smokers, alcoholics, cocaine and heroin addicts. Despite the fact that an abnormal sensitivity to delays controls the behavior of all of these patient groups, almost nothing is known about the mechanism of this disorder. We propose a series of experiments aimed at elucidating the basic mechanism that underlies the pathological decision making that marks inappropriate choice-in-time.

[unreadable] DESCRIPTION (provided by applicant): The past several decades have seen great strides in understanding the neuronal mechanisms of higher cognitive function in the intact brain, but research in this area is still seriously limited by the challenge of precisely targeting brain structures. Magnetic resonance imaging provides detailed information about brain structure, but this technology needs to be properly integrated with surgical procedures in the operating room to take full advantage of its power. In the clinical arena, functional neurosurgical procedures can be combined with MRI information using image-guided neurosurgical workstations. These instruments allow a surgeon to register surgical tools with a previously acquired MRI with high-resolution (<1 mm) and so precisely target any number of brain structures. The dramatic improvements in surgical outcomes and patient health mean this technology is now standard for work in humans. This proposal is for an image-guided neurosurgical workstation that will significantly increase the efficiency of on-going research at New York University into the brain mechanisms of higher cognitive functions. NYU has a significant number of investigators with research programs in this area studying visual development, perception, action, learning and memory in non-human primates. Understanding in this area will advance our ability to diagnose and treat mental health disorders, develop translational technologies such as cortical prostheses to help paralyzed patients, and understand disorders such as compulsive behavior and gambling. The instrument will be incorporated into the existing animal surgical facility at NYU and guide implantation of electrophysiology equipment necessary for this research. NYU already has a state-of-the-art Center for Brain Imaging that supports research into humans and animals. This instrument will leverage this existing strength to further research goals. Ultimately it will increase the productivity of current research, enable new research directions, and enhance animal health and safety. Relevance: Studying neuronal activity in the brain during cognitive processes like perception and movement planning is hard without a map of the brain because there are many small brain regions that perform these functions, and they can be in different places in different subjects. This proposal requests an advanced neurosurgical workstation that can guide studies of the brain by combining a map of the brain taken with magnetic resonance imaging with surgical tools during a surgery. This will let us quickly and easily find brain areas and investigate how they work together to give rise to cognition in ways that are currently unfeasible. [unreadable] [unreadable] [unreadable] [unreadable]

DESCRIPTION (provided by applicant): In the 1960s Barlow developed the efficient coding hypothesis, arguing that the function of sensory encoding is to maximize the efficiency with which neural signals are transmitted. "The principle of recoding [visual-sensory events into neural signals] is to find what images are expected on the basis of past experience and then to ... reserve the outputs with many impulses for the unusual or unexpected inputs." To guide visual behavior, movements of the eyes, efficiently encoded sensory signals in the striate and extra-striate cortex must be combined with stored estimates of the value, to the organism, of the actions associated with those signals. Perhaps surprisingly however, studies of the visual-saccadic system have lagged behind studies of the striate and extrastriate cortices in this regard. We know almost nothing about how reward-related saccadic signals are encoded, and this means that we have almost no understanding of how injuries and diseases that devastate brain areas ranging from the basal ganglia to the parietal cortex can be addressed by behavioral treatments. Here we propose to begin the formal study of neural coding in the saccadic system by programmatically assessing the efficiency with which the nervous system encodes information about the rewards that guide saccadic behavior in a way that will allow us to test several competing contemporary hypotheses about the function of the basal ganglia and the parietal cortex. We propose to test hypotheses about the roles of these brain areas in learning, storing, and representing the values of actions. Finally, we propose a set of causal interventions in brain activity to test the hypotheses developed by these behavioral and physiological experiments. We therefore propose to study 1) the midbrain dopamine neurons, 2) the visual-saccadic striatum (both the caudate and the ventral striatum) and 3) the lateral intra-parietal area. PUBLIC HEALTH RELEVANCE Patients with Parkinson's disease and other basal ganglia disorders are crippled by dysfunctional, or inefficient, movement generation. These patients also show aberrant reinforcement learning, although we do not now have a way of assessing the loss in learning efficiency that is associated with these diseases or how directly their deficit is related to learning. Relating activity in these areas to learning will allow us both to assess these deficits more accurately, and to relate measurable dysfunctions in learning to the well studied and crippling dysfunctions of movement that these patients suffer.

DESCRIPTION (provided by applicant): Our attitude towards risk shapes nearly every aspect of our behavior. It affects our willingness to take drugs of abuse, to make efficient financial investments, and to select amongst healthcare options. We know that these attitudes change over the lifespan and that these changes make us vulnerable in different ways as we age. Adolescents engage in risky behaviors that hugely increase their mortality rates. Elders typically show the opposite profile, avoiding risky behaviors to a fault. But what is it about our behavior that changes as we age? What specific behavioral propensities and neurobiological features of aging can account for the changing attitudes we take towards risks? In this application we propose to combine economic, psychological and neurobiological approaches to the study of risky behavior in an effort to paint a portrait of risk attitudes across the lifespan at the behavioral and neural levels. Behavioral economic analyses suggests that risk-taking behavior results from the three largely independent attitudes towards risks: (1) Technical risk-aversion, which reflects a trade- off between the probability of gains or losses and the magnitude of those gains and losses, (2) Ambiguity aversion, which reflects the subject's sensitivity to uncertainty, and (3) Loss aversion, which reflects a subject's differential sensitivity to losses and gains. We hypothesize that the reduction in risk-taking behavior across the lifespan that is widely observed is primarily due to an increase in ambiguity aversion, rather than to an increase in risk or loss attitudes. Further, we hypothesize that, if anything, humans should become less technically risk averse and loss averse as they age. At the neural level, we also know that there are clear structural patterns in brain development that may well underlie these behavioral changes observed across the lifespan. Of particular interest to our hypothesis is the observation that frontal regions mature late, actually thinning during development and actually continue to thin throughout the lifespan. Activational patterns fronto-cortical decision areas also change throughout the lifespan. These data suggest that both morphologic and functional properties of the frontal cortex may contribute to these changes in risk-related behavior over the lifespan, a hypothesis we propose to test with a mixture of behavioral and functional imaging techniques. PUBLIC HEALTH RELEVANCE: Our attitude towards risk shapes nearly every aspect of our behavior, including our willingness to take drugs of abuse, to make efficient financial investments, and to select amongst healthcare options. We know that these attitudes change over the lifespan and that these changes make us vulnerable in different ways as we age. We propose an analysis of the precise behavioral and neural basis of risk attitudes across the lifespan in an effort to better understand the neural basis of behaviors such as drug abuse, adolescent risk-taking, and healthcare decision-making.

DESCRIPTION (provided by applicant): In the 1960s Barlow developed the efficient coding hypothesis, arguing that the function of sensory encoding is to maximize the efficiency with which neural signals are transmitted. The principle of recoding [visual-sensory events into neural signals] is to find what images are expected on the basis of past experience and then to ... reserve the outputs with many impulses for the unusual or unexpected inputs. To guide visual behavior, movements of the eyes, efficiently encoded sensory signals in the striate and extra-striate cortex must be combined with stored estimates of the value, to the organism, of the actions associated with those signals. Perhaps surprisingly however, studies of the visual-saccadic system have lagged behind studies of the striate and extrastriate cortices in this regard. We know almost nothing about how reward-related saccadic signals are encoded, and this means that we have almost no understanding of how injuries and diseases that devastate brain areas ranging from the basal ganglia to the parietal cortex can be addressed by behavioral treatments. Here we propose to begin the formal study of neural coding in the saccadic system by programmatically assessing the efficiency with which the nervous system encodes information about the rewards that guide saccadic behavior in a way that will allow us to test several competing contemporary hypotheses about the function of the basal ganglia and the parietal cortex. We propose to test hypotheses about the roles of these brain areas in learning, storing, and representing the values of actions. Finally, we propose a set of causal interventions in brain activity to test the hypotheses developed by these behavioral and physiological experiments. We therefore propose to study 1) the midbrain dopamine neurons, 2) the visual-saccadic striatum (both the caudate and the ventral striatum) and 3) the lateral intra-parietal area. PUBLIC HEALTH RELEVANCE Patients with Parkinson's disease and other basal ganglia disorders are crippled by dysfunctional, or inefficient, movement generation. These patients also show aberrant reinforcement learning, although we do not now have a way of assessing the loss in learning efficiency that is associated with these diseases or how directly their deficit is related to learning. Relating activity in these areas to learning will allow us both to assess these deficits more accurately, and to relate measurable dysfunctions in learning to the well studied and crippling dysfunctions of movement that these patients suffer.

DESCRIPTION (provided by applicant): Our attitude towards risk shapes nearly every aspect of our behavior. It affects our willingness to take drugs of abuse, to make efficient financial investments, and to select amongst healthcare options. We know that these attitudes change over the lifespan and that these changes make us vulnerable in different ways as we age. Adolescents engage in risky behaviors that hugely increase their mortality rates. Elders typically show the opposite profile, avoiding risky behaviors to a fault. But what is it about our behavior that changes as we age? What specific behavioral propensities and neurobiological features of aging can account for the changing attitudes we take towards risks? In this application we propose to combine economic, psychological and neurobiological approaches to the study of risky behavior in an effort to paint a portrait of risk attitudes across the lifespan at the behavioral and neural levels. Behavioral economic analyses suggests that risk-taking behavior results from the three largely independent attitudes towards risks: (1) Technical risk-aversion, which reflects a trade- off between the probability of gains or losses and the magnitude of those gains and losses, (2) Ambiguity aversion, which reflects the subject's sensitivity to uncertainty, and (3) Loss aversion, which reflects a subject's differential sensitivity to losses and gains. We hypothesize that the reduction in risk-taking behavior across the lifespan that is widely observed is primarily due to an increase in ambiguity aversion, rather than to an increase in risk or loss attitudes. Further, we hypothesize that, if anything, humans should become less technically risk averse and loss averse as they age. At the neural level, we also know that there are clear structural patterns in brain development that may well underlie these behavioral changes observed across the lifespan. Of particular interest to our hypothesis is the observation that frontal regions mature late, actually thinning during development and actually continue to thin throughout the lifespan. Activational patterns fronto-cortical decision areas also change throughout the lifespan. These data suggest that both morphologic and functional properties of the frontal cortex may contribute to these changes in risk-related behavior over the lifespan, a hypothesis we propose to test with a mixture of behavioral and functional imaging techniques.

DESCRIPTION (provided by applicant): Existing computational models of dopamine's role in reinforcement learning are now quite well developed. These models make specific predictions about how changes in the firing rates of midbrain dopamine neurons should change the values subjects place on actions. While previous single neuron recording studies have largely validated these computational models, there have been very few efforts to use direct electrical manipulations of these neurons to examine these theories. This is of particular relevance because electrical manipulations of deep brain nuclei are now being used to treat a number of psychiatric disorders - including drug addiction. In this proposal, we describe a series of experiments aimed at testing the hypothesis that if the firing rates of midbrain dopamine neurons are manipulated with sub-second precision in a manner specified by existing computational theories, this may profoundly regulate behavioral preferences in a highly precise way. If this is the case, this finding would suggest a series of translationally-relevant experiments on the effects of deep brain stimulation on drug addiction. The current proposal seeks to lay the theoretical and experimental foundations for such experiments in the future.

DESCRIPTION (provided by applicant): Making decisions efficiently is the process of weighing costs and benefits. And while enormous progress has been made in the last two decades towards understanding the neural mechanisms by which humans and monkeys weigh benefits, much less is known about how the brains of these species weigh costs. There is, however, growing evidence from both humans and monkeys indicating that costs and benefits may be encoded by distinct neural systems and integrated quite late in the decision cascade. Here we propose a set of simple experiments designed to orthogonalize costs and benefits during human and monkey decision-making while we measure - and manipulate - neural activity in a suite of decision-related areas. Our goal is to test the hypothesis that separate anterior areas (mostly in the frontal cortex) maintain separate representations of costs and benefits, representations that are subsequently integrated for choice (in oculomotor tasks) in what are often called the fronto-parietal choice circuits. To test this hypothesis we propose a complementary set of functional Magnetic Resonance Imaging (fMRI) experiments in humans and single unit recording and muscimol inactivation experiments in monkeys. The fMRI experiments allow for a large-scale whole-brain test of our hypothesis. The single unit studies allow a rigorous quantitative measurement of the degree of cost and benefit representation in each area and the ability to establish a causal for some or all of these areas in cost benefit integration.

Project Summary/Abstract Opioid use disorder (OUD) is a debilitating chronic disease producing a growing burden on patients, providers, and the healthcare system. From 2002 to 2013, OUD rates have more than doubled and the number of individuals seeking treatment for the first time has more than quadrupled, driving unprecedented medical, scientific, and political interest in the etiology, pathophysiology, and treatment of OUD. Excellent treatments for opioid addiction exist, but their effectiveness is limited by lack of adherence to medication, treatment dropout, and relapse. There is scant knowledge about the neural and cognitive factors associated with and perhaps underlying treatment success or failure. A key goal of the present proposal is to develop reliable objective predictors of which individuals may need additional intervention and when best to intervene, i.e., when there is a risk for imminent relapse or treatment dropout. To do so, we propose to develop and test a computational neuroeconomic approach to quantifying the behavioral and neural features of addiction during OUD treatment. This computational approach to psychiatry seeks to understand circuit-level information processing in neural systems and how these mechanisms relate to normal and pathophysiological behavior. We hypothesize that quantifying individual subject choice behavior - via a longitudinally-sampled array of neuroeconomic decision tasks and models - provides information to: (1) distinguish relevant clinical populations (patients vs. controls and patient subgroups); (2) assist clinical prognosis (future treatment efficacy); (3) dynamically track ongoing clinical status (e.g. likelihood of relapse); and (4) examine the neural basis of behavioral changes in the recovery process. Specifically, we hypothesize that clinical status during treatment is characterized by the position and trajectory of individual subjects in a multidimensional space of decision parameters (quantifying impulsivity, risk tolerance, and ambiguity attitude). To test this hypothesis, we propose to longitudinally track the behavior and neural activity of patients seeking treatment for OUD. In Aim 1, we test the hypothesis that single-timepoint multidimensional decision data provides diagnostic and prognostic information, categorizing different clinical subpopulations (OUD patients vs. controls, treatment responsive vs. treatment refractory patients). In Aim 2, we test the hypothesis that dynamic multidimensional decision data tracks and predicts time-varying changes in clinical status, including the probability of future relapse. In Aim 3, we test the hypothesis that static and dynamic features of multidimensional decision data reflect corresponding features and changes in integrated value coding in specific neural circuits. Understanding how decision-related computations reflect clinical status is critical to closing the explanatory gap between biology and behavior in addiction. If successful, this approach offers both basic scientific and translational benefits: a clearer understanding of how and why behavior changes in addiction treatment, and easily-implementable tools to monitor treatment effectiveness and clinical course in individual patients.

Project Summary/Abstract Opioid use disorder (OUD) is a debilitating chronic disease producing a growing burden on patients, providers, and the healthcare system. From 2002 to 2013, OUD rates have more than doubled and the number of individuals seeking treatment for the first time has more than quadrupled, driving unprecedented medical, scientific, and political interest in the etiology, pathophysiology, and treatment of OUD. Excellent treatments for opioid addiction exist, but their effectiveness is limited by lack of adherence to medication, treatment dropout, and relapse. There is scant knowledge about the neural and cognitive factors associated with and perhaps underlying treatment success or failure. A key goal of the present proposal is to develop reliable objective predictors of which individuals may need additional intervention and when best to intervene, i.e., when there is a risk for imminent relapse or treatment dropout. To do so, we propose to develop and test a computational neuroeconomic approach to quantifying the behavioral and neural features of addiction during OUD treatment. This computational approach to psychiatry seeks to understand circuit-level information processing in neural systems and how these mechanisms relate to normal and pathophysiological behavior. We hypothesize that quantifying individual subject choice behavior - via a longitudinally-sampled array of neuroeconomic decision tasks and models - provides information to: (1) distinguish relevant clinical populations (patients vs. controls and patient subgroups); (2) assist clinical prognosis (future treatment efficacy); (3) dynamically track ongoing clinical status (e.g. likelihood of relapse); and (4) examine the neural basis of behavioral changes in the recovery process. Specifically, we hypothesize that clinical status during treatment is characterized by the position and trajectory of individual subjects in a multidimensional space of decision parameters (quantifying impulsivity, risk tolerance, and ambiguity attitude). To test this hypothesis, we propose to longitudinally track the behavior and neural activity of patients seeking treatment for OUD. In Aim 1, we test the hypothesis that single-timepoint multidimensional decision data provides diagnostic and prognostic information, categorizing different clinical subpopulations (OUD patients vs. controls, treatment responsive vs. treatment refractory patients). In Aim 2, we test the hypothesis that dynamic multidimensional decision data tracks and predicts time-varying changes in clinical status, including the probability of future relapse. In Aim 3, we test the hypothesis that static and dynamic features of multidimensional decision data reflect corresponding features and changes in integrated value coding in specific neural circuits. Understanding how decision-related computations reflect clinical status is critical to closing the explanatory gap between biology and behavior in addiction. If successful, this approach offers both basic scientific and translational benefits: a clearer understanding of how and why behavior changes in addiction treatment, and easily-implementable tools to monitor treatment effectiveness and clinical course in individual patients.

PROJECT SUMMARY Over 2 million Americans suffer from Opioid Use Disorder (OUD) and another 9 million misuse opioids. Treatments for opioid addiction exist, but effectiveness is compromised when subjects use illicit opiates during treatment. Reuse rates during treatment can be high, and reducing illicit opiate use during treatment has thus recently become a major NIDA policy goal. Elevated reuse rates not only compromise treatment effectiveness, but this behavior predicts, and likely drives, treatment dropout. With the support of a NIDA basic science R01, we developed a set of easy-to-use instruments that predict opioid reuse events with about twice the accuracy of any existing tool. The 5-minute battery we developed indicates the numerical probability that a patient will reuse illicit opiates within the next 7-10 days. In a pilot cohort, we successfully migrated this battery to a commercial smartphone platform, and demonstrated 100% retention and >85% compliance (median compliance > 95%) over a use period of up to 4 months. In a survey of our largely homeless MOUD patients we found that 85% already had smartphones and data contracts appropriate for using this platform as a part of their treatment. In a survey of OUD treatment physicians, we found that our system and the reuse prediction it provides, was both highly desirable and usable as implemented. Finally, we found in a reanalysis of data from CTN-0051 that dynamic dosing of this very kind reduces relapse rates. Our primary goal in this mid-scale clinical trial is to test the hypothesis that clinicians who use the output of our mobile system to adjust buprenorphine and methadone dosing achieve lower opiate reuse rates than physicians who provide care-as-usual. Our secondary goal is to examine the usability and desirability of this solution for clinicians with an eye to usability and large-scale deployment. Our third and final goal is to measure the cost-effectiveness of this solution from multiple perspectives. If we are successful it will be possible to employ an algorithmic and measurement-based approach to OUD treatment with methadone and buprenorphine which reduces reuse rates and relapse rates amongst OUD patients.