Modulating neuronal activity has implications for treating neurological disorders like Parkinson's disease, major depressive disorder, and drug-resistant epilepsy. Clinical neuromodulation is typically achieved by supplying electrical current to specific populations of neurons, thus triggering or inhibiting their activity. Apart from these clinical applications, advances in neuromodulation have enabled fundamental advances in our understanding of neural circuits. Some novel neuromodulation techniques that have emerged recently and enabled the elucidation of neural circuit function are optogenetics, chemogenetics, magnetogenetics, and sonogenetics.

Ultrasound neuromodulation and sonogenetics have the advantage of being non-invasive neuromodulation techniques. This method has a high spatial and temporal resolution in comparison to others being developed and this makes it an attractive option for eventual clinical translation. One of the key limitations in ultrasound neuromodulation is a lack of an understanding of the underlying mechanism of action of ultrasound on neurons. Specifically, it is unknown how mechanical energy carried by a sound wave is converted to the electrical activity of a neuron. An accurate understanding of the mechanism of action is critical to enable parameter selection and safety. The second key limitation in the field is the lack of devices specifically made for neuromodulation. Ultrasound has historically been used as an imaging modality but such devices are unsuitable for neuromodulation purposes. They cannot be driven at sufficient pressure without resulting in significant heating of tissue. This work seeks to address both of these limitations in the field. Intramembrane cavitation and lipid clustering have been proposed as likely mechanisms of ultrasound neuromodulation but they lack experimental evidence. We use high-speed digital holographic imaging to visualize membrane dynamics under the influence of ultrasound. We also develop a biophysical framework that couples ultrasound-induced pressure changes with a Hodgkin-Huxley neuron model to predict neuronal responses at relevant pressures. This work collectively demonstrates that membrane perturbations due to ultrasound result in the generation of action potentials. It provides a mechanism for both ultrasound-evoked neurostimulation and sonogenetic control.

We also address the phenomenon of standing acoustic waves in the skull producing large variations in pressure, leading to localized tissue damage and disruption of normal brain function. Our approach to overcome this involves the development of novel transducer-mounted diffusers that result in spatiotemporally incoherent ultrasound. We demonstrate the effectiveness of the diffuser both computationally and experimentally, and, show that the use of the diffuser results in a twofold increase in ultrasound responsiveness in cells with a sonogenetic candidate.

Finally, we demonstrate the development of tools for conducting ultrasound neuromodulation studies in vivo. These include a transparent ultrasound transducer that we use with a two-photon microscope to study ultrasound-induced activation of parvalbumin-positive neurons interneurons in the rodent visual cortex. We also develop a head-mounted transducer that can be combined with a fiber photometry system to stimulate activity in cholinergic neurons in the striatum.