Probing Phased-Array Focused Ultrasound Transducers Using Realistic 3D In-Silico Trabecular Skull Models: A Numerical Study

Transcranial focused ultrasound has emerged as a promising and innovative neuromodulation technique with the potential to foster brain disease therapies. Its efficacy depends on the precise targeting of specific brain regions using ultrasound waves. Achieving this precision requires the careful design of devices, with numerical simulations playing a critical role. Specifically, accurate modeling of the head structure is essential for optimizing the performance of the focused ultrasound transducer. This study investigates the impact of various skull models on the pressure field generated by two geometrically distinct focused ultrasound transducers. By quantifying the focal and scattered volumes beyond the skull, we gain valuable insights into the performance of each transducer. Furthermore, theanalysis emphasizes the importance of using a realistic porous skull model over a non-porous one inacoustic numerical simulations. The main goal of this work is to investigate numerical simulations outcomes using accurate skull models for testing numerically designed focused ultrasound transducers. Numerical simulations were performed using one homogeneous skull model (0% porosity) and two heterogeneous models with varying porosity levels (50% and 60%), that reflect the current understanding of human skull structure. The output pressure was generated by two distinct multi-element transducers (f-44 number = 0.8 and f-number = 1.1), each consisting of 96 active elements, positioned at a fixed distance above the skull models across all simulation scenarios. Acoustic numerical simulations were conducted using the MATLAB toolbox k-Wave and applying the time-reversal technique to correct for phase aberrations induced by the skull. The output data comprises 3D maps of pressure distribution. For each simulation, the Full Width at Half Maximum (FWHM) was calculated to obtain the -6 dB pressure isosurfaces, which were then used to determine the actual focal volume and the scattered pressure volume. Additionally, the focus size and its deviation from the target zone were analyzed to quantify the impact of the trabecular structure on ultrasound targeting. The numerical simulations revealed differences in the sonication performance of the two transducers. The non-porous model (0% porosity) exhibited almost no scattered volume compared to porous models (50% and 60% porosity) for both transducers. However, simulations using different porosity percentages did not show a consistent trend in variations of focus shift or FWHM. Neglecting the trabecular structure in transcranial numerical simulations results in inaccurate predictions of pressure distribution, potentially leading to unintended sonication of brain regions beyond the intended target. Simulating with homogeneous models can mask the occurrence of pressure hotspots and significant scattering beyond the skull, thereby creating a false impression of enhanced transducer performance.While the trabecular structure may not influence focus shift and size, it reveals a distinctly differentpressure distribution compared to homogeneous models. Consequently, employing heterogeneous models is crucial for accurately testing and predicting the pressure field generated by an ultrasound transducer. This approach ensures greater accuracy, precision, and selectivity, particularly when optimizing the transducer's configuration during the design and testing phases.