According to the World Health Organization (WHO), nearly 7.7 million new cases of dementia and Alzheimer’s are identified each year around the world. Moreover, the incidence of concussions in sports and their consequences for mental health have become of concern. Finally, despite a relatively low prevalence, brain cancers have a very poor prognosis, with a survival expectancy of less than one year for the most common and aggressive form, the glioblastoma multiform. Positron emission tomography (PET) is the imaging modality that is the most sensitive for studying non-invasively and in vivo these diseases at the biochemistry and cerebral metabolism levels, with the potential for earlier diagnosis, a more refined characterization of the type of pathology, but also a better understanding of the causes at the source of these diseases. Currently available PET scanners lack precision for detailed study of the brain. This project aims to develop a dedicated high performance PET scanner that will image the human brain with a spatial resolution 2 times higher than that reached with the best existing technology. At the planned resolution of 1.25 mm, it becomes possible to characterize tumor infiltration and tumor heterogeneity, determining factors of the evolution of aggressive cancers such as glioblastoma. It also becomes possible to follow in real-time osmotic opening of the blood-brain barrier to promote the transport of drugs to the tumor. Such a PET scanner will also allow molecular imaging of intermediate animals for biomedical research with unmatched precision.