Microglial cells have a considerable role in the pathophysiology of neuropsychiatric diseases. To study microglial cell function in vivo in clinical and translational settings, Positron Emission Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) imaging can be employed, using radiotracers binding specifically to the 18 kDa translocator protein (TSPO). This work describes multiple levels of study of TSPO as a microglial biomarker of in vivo brain imaging. First, the validation of a novel radiotracer of TSPO, [123I]CLINDE, as a biomarker for in vivo translational imaging is described, using a model of rat glioma. [123I]CLINDE was able to penetrate the blood brain barrier and to enter the brain parenchyma. It showed a high signal-to-noise ratio, and the majority of the radioactive signal was specifically bound to TSPO. Next, an innovative approach combining fluorescence activated cell sorting and ex vivo radioactive concentration measurement is introduced. The aim of this study was (i) to quantify which cell types in the brain, apart from microglia, contributed to the radioactive TSPO signal measured by molecular neuroimaging and (ii) to quantify which cell types contributed to the increase in TSPO concentration in animal models of brain disease and neuroinflammation. This study demonstrated that the TSPO radioactive signal is predominantly localized in microglia, although astrocytes and endothelial cells have a non-negligible contribution. In addition, the alteration of the TSPO radioactive signal concentration in disease has variable cellular origins, meaning that microglial, astrocytic and endothelial TSPO concentration alterations are disease dependent. The last work presented in this thesis described an in-depth assessment, in postmortem human brain tissue from a range of neurodegenerative diseases and their respective animal models, of the dynamics of TSPO expression in microglia. In this work, it was shown, using proteomic (immunohistochemistry, imaging mass cytometry), transcriptomic and genomic data that in human microglial cells, a proinflammatory response does not induce TSPO increase. TSPO concentration measured by molecular imaging in the human brain is thus indicative of the microglial cell density and not of proinflammatory activation in the human brain. Overall, this thesis describes the whole spectrum of considerations that must be addressed for the molecular neuroimaging of glial cell alterations in brain diseases.