The simulation of nonlinear propagation of ultrasound wave in biological tissue is a very time consuming operation. Different simulators, based on finite difference or angular spectrum methods have been reported in the literature and the second one provide faster simulations by considering separately the different harmonics. In this paper we proposed to use a generalized angular spectrum method (GASM) to quickly compute the first three harmonics of the nonlinear field. Moreover, the possible inhomogeneity of the medium is considered in the mathematical background. The GASM has been implemented on a graphic processing unit (GPU) in order to improve the computation time. The obtained results show a good agreement between the GASM and another finite difference nonlinear simulator. In terms of computation time, the GPU implementation allows a speed-up factor comprised between 3.5 and 13 depending on the GPU uses.
4D Simulation of nonlinear pressure field propagation on GPU with the Angular Spectrum Method / F. Varray; C. Cachard; P. Tortoli; O. Basset.. - ELETTRONICO. - (2011), pp. 250-253. (Intervento presentato al convegno IEEE International Ultrasonics Symposium tenutosi a Orlando, FL, USA nel Ottobre 2011) [10.1109/ULTSYM.2011.0061].
4D Simulation of nonlinear pressure field propagation on GPU with the Angular Spectrum Method
VARRAY, FRANCOIS;TORTOLI, PIERO;
2011
Abstract
The simulation of nonlinear propagation of ultrasound wave in biological tissue is a very time consuming operation. Different simulators, based on finite difference or angular spectrum methods have been reported in the literature and the second one provide faster simulations by considering separately the different harmonics. In this paper we proposed to use a generalized angular spectrum method (GASM) to quickly compute the first three harmonics of the nonlinear field. Moreover, the possible inhomogeneity of the medium is considered in the mathematical background. The GASM has been implemented on a graphic processing unit (GPU) in order to improve the computation time. The obtained results show a good agreement between the GASM and another finite difference nonlinear simulator. In terms of computation time, the GPU implementation allows a speed-up factor comprised between 3.5 and 13 depending on the GPU uses.I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.