Real-time ultrasound simulation using the GPU

In many ultrasound simulators for training purposes, synthetic ultrasound images have to be generated in real time. Patient specific anatomy can be modeled by using computed tomography (CT) data, and shadows can be calculated by using reflection coefficients and depth dependent, exponential attenuation. In order to include speckle, a pre-calculated texture map is typically added and blurred laterally to include an anisotropic point spread function (PSF). We propose to increase the speckle realism and allow for dynamic objects by using a physical model of the underlying scattering process. The scattering process is simulated as a convolution of the PSF of the ultrasound scanner with a scatterer distribution. The challenge is that the typical field-of-view (FOV) contains millions of scatterers which have to be selected by a virtual probe from an even larger body of scatterers. The main idea of this paper is to select and sample scatterers in parallel on the graphic processing unit (GPU). Ultrasound images of homogeneous tissue were produced in real time on a standard GPU (more than 10 frames per second). The ultrasound images were visually similar to images calculated by a reference method.