Functional ultrasound imaging (fUS) reports changes of blood volume with a mesoscopic spatial resolution. It is used to image brain activity. However, the extent to which the fUS signal quantitatively reports local brain activation is unknown. We propose to characterize the cellular and vascular signals underlying fUS responses to odor, and to quantitatively model the link between these signals in control and CADASIL mice. First, we will use the olfactory bulb as a neurovascular model and compare the fUS signal from a single voxel to cellular and hemodynamic signals evoked by odor, and measured with 2-photon microscopy in the same co-registered brain volume. We will then establish the transfer functions (TFs) linking these signals and test the TF robustness to predict fUS signals from cellular responses. Finally, we will use CADASIL mice to validate TFs as quantitative tools to follow vascular dysfunction with aging, and assess the deficit correction with immunotherapy.