RF-US thermoacoustic neuromodulation: a modeling effort
OA Version
Citation
Abstract
In recent decades, ultrasound has been investigated for neural modulation due to its
potential to provide high spatial resolution and low invasiveness. However, transcranial
application of ultrasound remains a challenge due to the high acoustic impedance of the
skull relative to the underlying neural tissue. This effect limits spatial resolution since
high frequencies are attenuated and reflected within bone, allowing only low frequencies to
effectively couple into neural tissue. To overcome these limitations, a thermoacoustic approach is considered in which RF is used to transmit signal past the skull and better couple high frequency ultrasound to the brain. Ultrasound signal could potentially be superimposed and steered within the brain through the use of a phased array of RF antennas placed
outside the skull. To investigate this effect, this thesis presents the development of a 2D
finite difference code that incorporates varying RF input parameters to simulate ultrasound
initial pressure, velocity, and attenuation through layered phantom and human tissues. In
particular, the simulator accounts for RF absorption and fluence within different materials
due to varying antenna specifications, including peak power, pulse width, antenna gain, and
aperture size. Results of various simulations are discussed, suggesting that increasing the
number of antennas in the RF array allows for increased focusing and higher intensity of
signal at a distance from the source than can be achieved by one antenna alone. Safety considerations are modeled to determine if adequate ultrasound intensity of 0.3-0.8 MPa can be provided through the thermoacoustic effect to modulate neural activity, while remaining below IEEE safety exposure guidelines of 20 mJ/cm2. Further investigation of more complex array configurations is required to determine if energy density can be spread over a large enough area to lower exposure risk, while providing clinically relevant pressures and frequencies for the purpose of neural modulation. Finally, the finite difference code is used to verify results from a preliminary thermoacoustic array design using omnidirectional antennas that was performed at MIT Lincoln Laboratory. It was found that antennas
with larger gain will provide more efficient conversion of RF to ultrasound.