Comparison between ray-tracing and full-wave simulation for transcranial ultrasound focusing on a clinical system using the transfer matrix formalism.

Only one High Intensity Focused Ultrasound device has been clinically approved for transcranial brain surgery at the time of writing. The device operates within 650 kHz and 720 kHz and corrects the phase distortions induced by the skull of each patient using a multi-element phased array. Phase correction is estimated adaptively using a proprietary algorithm based on computed-tomography (CT) images of the patient's skull. In this paper, we assess the performance of the phase correction computed by the clinical device and compare it to (i) the correction obtained with a previously validated full-wave simulation algorithm using an open-source pseudo-spectral toolbox and (ii) a hydrophone-based correction performed invasively to measure the aberrations induced by the skull at 650 kHz. For the full-wave simulation, three different mappings between CT Hounsfield units and the longitudinal speed of sound inside the skull were tested. All methods are compared with the exact same setup thanks to transfer matrices acquired with the clinical system for N=5 skulls and T=2 different targets for each skull. We show that the clinical ray-tracing software and the full-wave simulation restore respectively 8415% and 8615% of the pressure obtained with hydrophone-based correction for targets located in central brain regions. On the second target (off-center), we also report that the performance of both algorithms degrades when the average incident angles of the acoustic beam at the skull surface increases. When incident angles are higher than 200, the restored pressure drops below 75% of the pressure restored with hydrophone-based correction.