Despeckling of medical ultrasound images using fast bilateral filter and neighShrinkSure filter in wavelet domain

Abstract

The diagnostic quality of medical ultrasound (US) images is affected by a multiplicative type of noise known as speckle noise. In this paper, a new denoising scheme based on thresholding of wavelet coefficients in different sub-bands is presented. NeighShrinkSure filter is used for thresholding detail band wavelet coefficients (high pass component). Also, as in medical US images approximation band coefficients (low pass component) also consist of speckle noise so fast bilateral is applied on these coefficients to improve the performance of proposed method. Experiments were performed on synthetic and real US images. The performance of the proposed method with four other existing methods is evaluated objectively and subjectively. Objective evaluation is carried out using parameters PSNR, SNR, SSIM and FOM and for subjective evaluation denoised US images obtained from all methods are inspected visually. The results obtained illustrate the effectiveness of the proposed method over other existing methods.