Machine learning-based optimization of a single-element transcranial focused ultrasound transducer for deep brain neuromodulation in mice

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Transcranial focused ultrasound is an emerging noninvasive technique for neuromodulation, offering high spatial precision and potential for deep brain penetration. For ultrasound neuromodulation studies in mice, current transducers can successfully reach shallow and middle brain depths; however, accessing deep-brain regions, especially those near the lower skull bone, remains challenging due to skull-induced attenuation, reflection, and acoustic aberration. To address this issue, this study presents a machine learning-based computational framework for optimizing single-element transducer designs to achieve precise deep-brain targeting in a mouse model. This framework includes a surrogate model and an evolutionary optimization approach. The surrogate model, which consists of Random Forest regressor and classifier models, was trained on acoustic simulation results to predict acoustic performance from design parameters. A total of 72 simulations were performed across coronal and sagittal planes, systematically varying parameters of frequency (1-6 MHz), radius of curvature (5-7 mm), and f-number (0.58-1.0). Each design was evaluated using five performance metrics: focal length, focal shape, maximum pressure at the focal region, pressure maximum location, and sidelobe suppression. The surrogate was combined with an evolutionary algorithm, Non-Dominated Sorting Genetic Algorithm II, to perform multi-objective optimization and identify high-performing transducer designs. The optimized design resulted in more compact, symmetric focal regions and accurate energy delivery to deep targets, with minimal off-target exposure, even in the presence of complex skull anatomy. Results indicate that lower f-numbers, moderate radius of curvature, and higher frequencies facilitate precise deep brain targeting. Overall, this data-driven approach enables the effective design of single-element transducers for deep-brain neuromodulation in mice and provides a framework for designing transcranial transducers for other brain targets, potentially accelerating the clinical translation of focused ultrasound technologies.

Article activity feed