Synthetic aperture ultrasound imaging using GPUs

Abstract

Ultrasound images are widely used for medical diagnosis and for Non-Destructive Testing (NDT) in the industry. Among alternative imaging methods there are the Phased Array techniques, which allow processing ultrasound signals in real time and generating focused images within certain limits. On the other hand, there are synthetic aperture techniques, commonly called SAFT, which are post-processing procedures that allow the focusing of all image points in both emission and reception. However, the use of SAFT techniques often are conditioned on industry by the high computational cost that require. This paper deals with the implementation of an algorithm for synthetic aperture imaging using a parallel processing platform. The development has been tested and validated on reference parts and nuclear industry testing pieces.