Performance of Adaptive Beamformers for Ultrasound Imaging of a Partially Shaded Object

Ultrasound imaging of tissue involving bone structures are prone to poor image quality due to different artifacts. The shadowing effect owing to obstruction of the ultrasound beams leads to reduced resolution and in some cases geometrical distortion. Minimum variance beamforming method (MV) using spatial averaging technique for the covariance matrix estimation has been successfully applied to medical ultrasound imaging. It has showen significant improvements in the image quality when the full imaging aperture exists. We demonstrate that the robustness of MV beamforming degrades when the aperture is highly obstructed by an acoustically hard tissue. In a simulation and an in-vitro study, we show that this robustness deficiency can distort image of the point scatterers located in the shadow of a bone tissue. This distortion can be seen as a lateral shift of the point spread function (PSF) and a decrease of the sensitivity. We show that forward/backward (FB) averaging of the covariance matrix increases the robustness of the MV beamformer against the shadowing at expense of the resolution degradation. Employing FB averaging technique, together with the eigenspace minimum variance (ESMV) beamformer can improve the performance of the MV beamformer.