Abstract TP253: Hybrid Angular Spectrum-based Acoustic Window Modeling for Rapid Transcranial Ultrasound Stroke Detection

The Hybrid Angular Spectrum (HAS) method can be used to calculate ultrasound wave propagation through inhomogeneous tissue geometries, providing fast and accurate results. Volumetric matrices of sound speed and attenuation and a planar incident pressure pattern are given as inputs, and the pressure wave propagation pattern is calculated plane-by-plane, by alternating between the space domain and the spatial-frequency domain for each slice of the volume. In this work, the HAS method was implemented in MATLAB to reconstruct ultrasound fields through inhomogeneous 3D models of the temporal skull. Accuracy of the HAS code was validated by comparison of pressure field distributions simulated in a homogeneous water medium with results from Field II. The 3D spatial gradient of the phase-shift error between forward- and back-propagated fields was calculated and used to adjust sound speed values within a model of the skull. Multiple iterations of this process led to convergence to accurate sound speed values within 3 m/s. CT scans of ex vivo human skulls were used to run simulations through sections of temporal bone. The resulting digital acoustic models of temporal skull will be used to inform custom array probe design to help identify ideal acoustic windows for transcranial ultrasound acquisition and streamline LVO detection by paramedics.