Project Summary/Abstract The current opioid epidemic is a pressing public health crisis. A key precipitating factor of reuse and relapse among people with opioid use disorders (OUD) is craving, or the intense, specific desire for the drug. While craving has been extensively studied, and is known to predict drug use, we still lack an explanatory and algorithmically-precise model that can directly link craving neurobiology to its observed consequences: the decision to pursue drugs over other valuable alternatives. Given that typical treatments for OUD do not adequately address craving and fail to prevent reuse in many patients, clarifying the precise, decision-relevant, mechanism of craving may critically inform more targeted ways to treat craving and improve clinical outcome. To address these important questions, we developed an experimental paradigm to study craving based on methods widely used in decision neuroscience to assess value-based decision-making. Decision neuroscience (or neuroeconomics) integrates concepts and methods from psychology, economics, and neuroscience to understand the neural architecture for decision-making, and has been increasingly applied in mechanistic studies of psychiatric disorders including addiction. Our paradigm constitutes a novel application of this framework by quantifying a subject’s in-the-moment (i.e., state-dependent) decision process during craving15. In pilot behavioral studies in healthy and opioid addicted subjects, we find that this paradigm captures 1) how value—the key determinant of the decision to pursue a particular option versus another—changes under craving, and 2) the selectivity of this effect to the object of craving. It also 3) provides an algorithmically-specific process (a mathematical description) of this change that can be used to tie behavior to its neural substrate. In the present study we aim to elucidate this neural substrate by identifying the specific neural computations through which craving modulates the value of drug and nondrug alternatives and thereby drug use decisions in human OUD. We propose to identify the neural substrate of opioid craving in N=89 OUD patients who will complete our paradigm during fMRI in a within-subjects cross-over design following a brief craving induction or a control manipulation16. Because decision circuits encode value in a reward-identity specific manner, our design will enable us to isolate the computations associated with drug-related value from those of nondrug value. Our study will for the first time determine whether and how experimentally-induced craving dynamically shifts such “identity-specific” neural encoding of drug-related value (Aim 1), and the parts of a putative ‘craving circuit’ involved in this shift (Aim 2). To test whether this mechanism is unique and reward-identity specific, we will also measure brain activity associated with experimentally-induced food craving and specific food-value in the same patients and N=89 healthy con...