Publication Details
Acceleration of Ultrasound Neurostimulation Using Mixed-Precision Arithmetic
Duchoň Radek, Ing. (FIT BUT)
GPU, Nvidia, CUDA, k-Wave, Acceleration, Ultrasound, Acoustic
waves, Neurostimulation, Mixed precision
Ultrasound neurostimulation, a technique that modulates the brain's electrical activity, has emerged as a significant secondary treatment option for cases resistant to pharmacological interventions. The therapy is achievable through the application of a three-dimensional steerable ultrasound, directed by patient-specific stimulation plans. These plans are meticulously crafted through full-wave ultrasound propagation simulations. Nonetheless, the computational intensity required for calculating these plans poses a significant challenge, often reaching the memory capacities of contemporary graphics processing units (GPUs). By representing material properties and k-space operators more efficiently, we achieved a 22% reduction in precision GPU memory usage, while accelerating calculations by 8.5%. This optimization introduced an error that reduced focal pressure by 0.5% without any focus movement, values that are clinically acceptable.
@INPROCEEDINGS{FITPUB13194, author = "Ji\v{r}\'{i} Jaro\v{s} and Radek Ducho\v{n}", title = "Acceleration of Ultrasound Neurostimulation Using Mixed-Precision Arithmetic", pages = "370--372", booktitle = "HPDC '24: Proceedings of the 33rd International Symposium on High-Performance Parallel and Distributed Computing", year = 2024, location = "New York, US", publisher = "Association for Computing Machinery", ISBN = "979-8-4007-0413-0", doi = "10.1145/3625549.3658823", language = "english", url = "https://www.fit.vut.cz/research/publication/13194" }