High-contrast and low-computational complexity medical ultrasound imaging using beamspace capon method.

Abstract

Several adaptive beamforming techniques have been proposed to improve the quality of medical ultrasound images. The beamspace (BS) Capon method is one common method used to depict high-resolution images with low computational complexity. However, the complexity is not low enough for real-time imaging in clinical situations because the conventional BS Capon method employs a time-delay process and a transition process from elementspace signal processing to BS signal processing at all points of interest. Thus, we propose a technique that replaces the time-delay process using a steering vector. In addition, the Capon method employs a spatial averaging (SA) technique to stabilize the estimation in intensity. However, when the averaging size is not adequate, the estimated intensity might be smaller than that given by the delay-and-sum (DAS) method. Because most medical diagnoses are presented based on the estimation of intensity acquired by the DAS beamformer, accurate estimation of intensity is also required. Therefore, we propose a compensation technique that uses both small and large sizes for SA. In an experiment, the -6 dB beam width, sidelobe level, and estimation error in the intensity of the proposed method were 0.17 mm, -27 dB, and 0.92 dB, respectively, where those of the conventional BS Capon method were 0.29 mm, -22 dB, and 8.1 dB. The complexity of the proposed method is one-fourteenth that of the conventional method. Compared with conventional methods, the proposed method succeeded in depicting a higher-contrast image with accurate intensity estimation and lower computational complexity.