Brain function is mediated by hierarchical local and long-range circuits organized across multiple spatial scales. Bridging and spanning these scales of organization is essential for understanding brain function and ultimately dysfunction; however, no single existing technology can accomplish this daunting task. Human brain activity and connectivity can be studied with non-invasive magnetic resonance (MR) methods that can cover the entire brain. However, the spatiotemporal resolution and fidelity to neuronal activity is limited because of the intervening neurovascular coupling that is the source of the MR mapping signals. These limitations can be overcome if MR resolutions and fidelity can be improved so as to reduce the heterogeneity of the responses within an MR voxel and, in addition, the MR method is combined with other techniques that simultaneously report on neuronal and/or neurovascular responses, ideally sampling the activity within one or more MR voxels at the single neuron, synapse or vessel level. Besides interrogating the link between neuronal activity and the MR based functional mapping signals, the complementary nature of such a combination of techniques would provide the means for bridging the multiple spatial and temporal scales, going from the cellular and synaptic level to coordinated activity over billions of neurons spanning large parts of the brain, if not the entire brain. This TRD approaches this problem by proposing to develop i) advanced MR methods for imaging brain function and connectivity at unprecedented spatial resolution using very high magnetic fields, ii) combining such MR measurements with simultaneous measurements of multi-photon recordings of neural signals (at single cell and/or synapse level) and corresponding hemodynamic responses at the level of individual arterioles, capillaries and venules, within the environment of an ultrahigh field (UHF) magnet on the same animal and under the same experimental conditions. Because of the invasive nature of the optical methods the combined experiments can only be performed in animal models while the MR techniques to be developed would be applicable to human imaging as well.