Stroke is a devastating disease. For ischemic stroke, recent advances with endovascular thrombectomy (EVT) enable revascularization of the occluded artery in a fast and efficient way. Despite this revolutionary intervention and successful recanalization, over half of patients remain disabled. Current imaging fails to reliably predict this outcome. Preserving regional cerebral metabolism is essential to limit the damage. Positron Emission Tomography (PET) can measure cerebral metabolic rate of oxygen consumption (CMRO2) and oxygen extraction fraction (OEF). Indeed, the concept of penumbra in ischemic stroke came from PET CMRO2, but this reference method for the biomarker is incompatible with acute stroke care. CMRO2 can also be evaluated using magnetic resonance imaging (MRI), but the MRI method still suffer from major limitations, including the lack of a fast and robust post-processing method. We propose to validate this new biomarker together with a clinically usable software using our unique preclinical model in non-human primates (NHP). Our ambition is to transfer to clinical application for a first in-man clinical trial in the context of EVT. The operational organization within our consortium will ensure an effective management (WP1). We will first take advantage of our unique database (ANR CYCLOPS, RHU Marvelous) with 23 NHP longitudinal CMRO2 PET-MRI at baseline, during occlusion and follow-up up to 30 days post-EVT. We will use these data to build a dictionary of the MR parameter kinetics to propose a robust post-processing pipeline based on regional vascular specificities (WP2). The retrospective study will analyze the different brain region patterns longitudinally to distinguish between degenerative and repair/compensatory processes, potential key factors of clinical outcome (WP 3.1). In the prospective phase, NHP will be treated during occlusion to improve their collateral status. For the first time, we will use MRI CMRO2 as a companion biomarker for an innovative non-pharmacological neuroprotective intervention (WP 3.2). Indeed, MRI CMRO2 will be used to evaluate the treatment effect both at the acute and subacute phases and will be optimized for clinical transfer. We will then perform a clinical trial to acquire post-EVT PET-MRI clinical data in 20 patients at the subacute phase (WP4). If acquisition time optimization is efficient, a clinical trial at the acute phase could be planned rapidly. A prototype of a robust post-processing pipeline implemented in Python will be tested using the retrospective preclinical data and published clinical data to validate its efficiency (WP5.1). The module will then be implemented in the Olea Sphere® clinical platform for cerebrovascular diseases (WP5.2). Its reliability, robustness and diagnostic efficiency will be evaluated in the preclinical and clinical protocols (WP3 & WP4) for future deployment at stroke acute phase. The consortium associates internationally recognized preclinical and clinical expertise and distinctive MRI and post-processing competencies in a public/private partnership. Our project will make CMRO2 measurements available for routine clinical use, including acute stroke care. Multiple MRI research fields may benefit from this development, including cerebrovascular physiological studies.