*Context* Reward is essential for motivation and learning, but can also lead to maladaptive behaviors such as addiction. Reward processing in the brain has been described at various levels, with a large part of the animal literature focusing on subcortical dopamine, while the human literature has relied more strongly on blood oxygen level dependent (BOLD) fMRI. An important challenge is to integrate these levels of description, in order to provide a more unified picture of the healthy and dysfunctional mechanisms underlying reward processing in the brain. In addition, the recent advent of dimensional psychiatry has emphasized the multi-factorial nature of psychiatric disorders and the importance of going beyond single cognitive dysfunctions –such as reward processing– in order to improve the current nosography and treatment strategies. *Objectives and methodology* This project is at the crossroads of fundamental and clinical neuroscience, pursuing two main goals. First, we will attempt to bridge molecular and functional levels of description of the brain mechanisms supporting reward processing in healthy humans. Specifically, we will address the question of how reward-induced dopamine release in the striatum is (i) regulated by top-down prefrontal control and (ii) coupled with striatal BOLD activity. We will address these questions using a state-of-the-art technique –simultaneous PET-fMRI– combined with a machine learning approach and the novel ‘lp-ntPET’ analytical framework optimized for tracking dynamic changes in dopamine release (Axis 1). Then, building on this knowledge and multi-modal approach, we will focus on gambling addiction, a psychiatric disorder that has been associated with impaired reward-related dopamine functioning. Despite the popularity of this hypothesis, supporting evidence remains inconsistent. To resolve these inconsistencies, we propose to test a novel model of gambling addiction which (i) explicitly takes into account individual differences, and (ii) posits an interaction between abnormal reward processing and threat reactivity in the brain. Importantly, we will test the ecological validity of this neurobiological model, by examining its predictive value for real-life gambling behavior using Ecological Momentary Assessment (Axis 2). *Impact* This project has the potential to make an impact at three levels. At a fundamental level, this project will provide a multi-level description of reward processing in the brain. This should help refine models that relate striatal dopaminergic transmission with hemodynamic changes in the human brain, and inform our understanding of how top-down prefrontal control of reward behaviors is implemented within cortico-striatal loops. At a methodological level, the combined use of simultaneous PET-fMRI and machine learning should allow us to derive a proxy measure of striatal dopamine release based on BOLD fMRI. Given that fMRI is much cheaper and less invasive than PET, such a proxy measure would be highly valuable for assessing subcortical dopamine dysfunctions across a range of psychiatric disorders, such as addiction, depression or schizophrenia. Finally, at the clinical level, this project should bring the first proof-of-concept that a model of gambling addiction explicitly accounting for individual differences and a multi-factorial etiology can reconcile previous inconsistent findings, and make predictions about real-world behavior. This approach relying on cognitive dimensions, as opposed to rigid categorical diagnoses, should eventually contribute an improved biological definition of addiction, and pave the way to more personalized treatment strategies.

DAVincCCHeaP answers research issues about clean and sustainable energies as well as reduction of primary energy consumption. Combining theoretical, numerical (CFD) and experimental approaches, this project will support French innovations in the field of regenerative thermal machines. There is great interest in intelligent equipment that supplies several energy sources simultaneously. Scientific issues are addressed through a unique test bench whose data will feed a predictive modelling tool allowing the development of such machines. Heat transfer, multi-scale physics and innovative thermodynamic cycles are the main keys to improve the performance and competitiveness of industrial equipment. Technological advances will benefit to a large range of applications from domestic heat pumps and engines to space cryocoolers. This project with high dissemination potential will promote technological innovation and address the energy and environmental challenges of tomorrow's society.