This proposal responds to BRAIN Initiative RFA-NS-20-006 and aims to elucidate the neural and biophysical mechanisms of noninvasive focused ultrasound (FUS) neuromodulation. FUS overcomes shortcomings of other neuromodulation methods and can noninvasively stimulate millimeter-scale regions in any part of the brain including deep brain structures. We seek to understand how different doses and spatiotemporal applications of FUS interact with the brain at cellular, circuit, and behavioral level. When used in conjunction with MRI, the FUS beam can be precisely localized while network-level effects can be observed with BOLD fMRI. In the past few years, through BRAIN Initiative funded projects, we have developed an integrated MRI guided FUS system (MRgFUS) with image-guidance and MRI capabilities required to place the beam accurately in the brain and map its location using magnetic resonance acoustic radiation force imaging (MR- ARFI). Using this system, we have demonstrated that FUS exerts bidirectional (excitatory and inhibitory) and state dependent neuromodulation of the nonhuman primate (NHP) sensorimotor system. FUS directly excites somatosensory area 3a/3b neurons at resting state but suppresses activated neurons when they are engaged in processing tactile inputs and elicits activation in downstream off-target brain regions. Here, we seek to investigate the mechanisms underlying FUS neuromodulation by evaluating neural signals at multiple scales using multiunit array electrodes and functional MRI during FUS neuromodulation over a parameter space chosen to test the influence of pulse duration, pulse repetition frequency, and amplitude. The planned studies will use optogenetics and pharmacological manipulations to test the hypothesis that increasing repetition frequency independent from other parameters preferentially drives specific groups of neurons. Studies varying amplitude will assess a hypothesis derived from our recent observation that FUS at moderate pressures elicits stronger inhibitory effects than high pressures. We will map invasive electrophysiological measurements to non-invasive measurements of neural activity (e.g. BOLD fMRI) that can be used in humans. Safety will be assessed with imaging, deep learning analysis of hand grasping behavior, and post-mortem assessment, providing important information for the ongoing translation of FUS neuromodulation. Our proposed studies will elucidate mechanisms underlying FUS neuromodulation over a broad parameter space in experimental models that span the individual neurons through whole brain networks and connect these multi scale electrophysiological and functional MRI observations made at the cellular, local microcircuit, and global levels to behavior changes in a NHP model system. Thus, our approach is closely aligned with the goals of this RFA and will further our knowledge of FUS neuromodulation, a fast-growing non-invasive method for dissecting circuits in the mammalian brain that offers th...