Adaptive Diagonal Reducing and Adaptive Weighting Approach to Covariance Matrix-Based Statistical Beamforming for Ultrasound Imaging

We previously proposed the covariance matrix-based statistical beamforming (CMSB) for ultrasound image reconstruction, and it has been demonstrated obviously improved resolution and contrast over delay-and-sum (DAS). However, the speckle smoothness obtained by CMSB is inadequate in phantom experiment, and the constant diagonal reducing factor offers a tradeoff between resolution, contrast, and speckle quality. This paper aims to improve the speckle quality of CMSB and overcome the tradeoff offer by the constant diagonal reducing factor. We introduced adaptive diagonal reducing and adaptive weighting for CMSB to improve speckle quality and contrast, respectively. Furthermore, covariance matrix with a size threshold and a tanh function as adaptive weighting for CMSB to further preserve speckle signals and reduce off-axis clutter, respectively. We used a simulated and a phantom experimental datasets to evaluate the imaging performance of the proposed method. Results show that the proposed method significantly improves the imaging quality of the previously proposed CMSB.