Equipment and methods for synthetic aperture anatomic and flow imaging

Abstract

Conventional ultrasound imaging is done by sequentially probing in each image direction. The frame rate is, thus, limited by the speed of sound and the number of lines necessary to form an image. This is especially limiting in flow imaging, since multiple lines are used for flow estimation. Another problem is that each receiving transducer element must be connected to a receiver, which makes the expansion of the number of receive channels expensive. Synthetic aperture (SA) imaging is a radical change from the sequential image formation. Here ultrasound is emitted in all directions and the image is formed in all directions simultaneously over a number of acquisitions. SA images can therefore be perfectly focused in both transmit and receive for all depths, thus significantly improving image quality. A further advantage is that very fast imaging can be done, since only a few emissions are needed for forming an image, and a novel approach of recursive ultrasound imaging can be used to give several thousand images a second. A commercial SA imaging system has, however, not yet been introduced due to a number of problems. The fundamental problems are primarily that the signal-to-noise ratio and penetration depth are low and velocity imaging is thought not to be possible. This paper will address all the issues above and show that they can all be solved using various techniques. The SNR is increased significantly beyond that for normal systems by using coded imaging and grouping of elements to form larger defocused emitting apertures. It is also possible to have many more receive channels, since different elements can be sampled during different emissions. The paper also shows that velocity imaging can be performed by making a special grouping of the received signals without motion compensation by using recursive imaging. With this technique continuous imaging at all points in the image is possible, which can significantly improve velocity estimates, since the estimates can be formed from a large number of emissions (100-200). The research scanner RASMUS, capable of acquiring clinical SA images, has been constructed and will be described. A number of phantom and in-vivo images will be presented showing in-vivo SA B-mode and flow imaging.