Nonlinear Diffusion based on Bayesian Estimator in Laplacian Pyramid Domain for Ultrasonic Speckle Reduction

Abstract

Laplacian pyramid based nonlinear diffusion (LPND) method is proposed for speckle noise reduction in medical ultrasound imaging. In this method, speckle is removed by nonlinear diffusion filtering of bandpass ultrasound images in Laplacian pyramid domain. For nonlinear diffusion in each pyramid layer, a Bayesian threshold is automatically determined by a variation of robust median estimator. The proposed LPND method reduces speckle noise significantly while preserving sharp image details. The superiority of the proposed LPND over the conventional LPND based on Gaussian filtering and LPND method based on Butterworth filtering based speckle reduction technique is demonstrated by applying them on synthetic and real ultrasound B-mode images.