Modeling of transcranial ultrasound for therapeutic and diagnostic applications

Abstract

Ultrasound’s use in the brain has conventionally been limited by its inability to penetrate the skull. To overcome these limits, we have been investigating techniques to maximize energy transfer and minimize distortion through the skull bone. These model-based aberration correction approaches - now in the early stages of clinical testing - rely on both practical and accurate numeric methods. Efforts to improve these methods necessitate an increasingly detailed consideration of skull heterogeneity. To facilitate this numerically-intensive problem, we are utilizing an inhomogeneous pressure simulation code, based on a pseudo-spectral solution of the linearized wave equation. Forward and scattered waves are determined over a pre-specified volume with scattering determined by the impedance mismatch between a given voxel and regional points in the projection plane. The total forward-scattered pressure is recorded over the relevant k-space, while reflected energy is processed in a separate backward projection. This process is repeated iteratively along the forward projection plane until the volume of interest has been traversed. This procedure can be repeated an arbitrary number of times N, representing N-1 order scattering. Abilities and limitations of the method will be demonstrated by comparison with FDTD simulation.