Research On Adaptive Beamforming Approaches In Ultrasound Imaging

Beamforming is one of important components in medical ultrasound imagingsystems. Its performance has a direct impact on the quality of ultrasound image.Currently, the most commonly used beamforming algorithm is the delay-and-sumbeamforming, which forms an ultrasonic beam with high sidelobes. Usually, thesidelobe level can be suppressed by using amplitude apodization, yet it will result inthe increase of the mainlobe width. Unlike non-adaptive beamforming, adaptivebeamforming calculates the weights based on the ultrasound field. Therefore, thisdissertation focuses on the adaptive beamforming applied in ultrasound imaging.Firstly, the basic theory of the delay-and-sum beamforming is introduced, andthen some beam control methods are analysed, such as amplitude apodization anddynamic focusing. However, it is impractical to implement real dynamic amplitudeapodization, and the dynamic focusing can not be applied to transmission of beams.Then, the Capon adaptive beamforming algorithm is introduced to implementthe real dynamic apodization. In this way, the dynamic weights are calculated usingthe characteristics of the echo signals. In order to obtain a better estimate of thecovariance matrix which is employed in adaptive weights calculation, theforward-backward spatial smoothing technique is used to decorrelate the coherentsignals received by transducer elements. Furthermore, the diagonal loadingtechnique is applied to improve the robustness of the Capon adaptive beamformer,and also this dissertation discusses how to set the diagonal loading parameter.Because the dynamic focusing can not be used in transmission in the delay-and-sum beamforming, several scanning modes of synthetic aperture are introduced toimprove the quality of ultrasound images in this dissertation. And then themulti-element synthetic aperture focusing mode is applied in adaptive beamformingto obtain the echo signals, which is then processed by the Capon algorithm once andtwice respectively. The simulation shows that the image quality of the latter is better.Finally, due to the large amount of computation, the Capon adaptivebeamforming algorithm is not suitable for real-time imaging, therefore, a simplifiedscheme is proposed. The computation analysis and simulations show that thesimplified scheme, to some extent, reduces the resolution and contrast of ultrasoundimages. However, the amount of computation of the Capon algorithm is reducedeffectively, and the running speed of the program is enhanced.