An improved adaptive wavelet shrinkage for ultrasound despeckling

Abstract

Ultrasound imaging is the most widely used medical diagnostic technique for clinical decision making, due to its ability to make real time imaging for moving structures, low cost and safety. However, its usefulness is degraded by the presence of signal dependent speckle noise. Several wavelet-based denoising schemes have been reported in the literature for the removal of speckle noise. This study proposes a new and improved adaptive wavelet shrinkage in the translational invariant domain. It exploits the knowledge of the correlation of the wavelet coefficients within and across the resolution scales. A preliminary coefficient classification representing useful image information and noise is performed with a novel inter-scale dependency measure. The spatial context adaptation of the wavelet coefficients within a subband is achieved by a local spatial adaptivity indicator, determined by using a truncation threshold. A weighted signal variance is estimated based on this measure and used in the determination of a subband adaptive threshold. The proposed thresholding function aims to reduce the fixed bias of the soft thresholding approach. Experiments conducted with the proposed filter are compared with the existing filtering algorithms in terms of Peak-Signal to Noise Ratio (PSNR), Mean Square Error (MSE), Structural Similarity Index Measure (SSIM), Equivalent Number of Looks (ENL) and Edge Preservation Index (EPI). A comparison of the results shows that the proposed filter achieves an improvement in terms of quantitative measures and in terms of visual quality of the images.

RETRACTED: An efficient approach for brain tumor detection and segmentation in MR brain images using random forest classifier

“…Programmed brain extraction strategies was proposed for T1 magnetic resonance pictures utilizing area based marking and morphological activities (Chinnathambi et al, 2014; Devi and Asokan, 2014). Initially, the binary image was obtained from MR images by an adaptive intensity thresholding technique which was unsupervised and knowledge based perception.…”

Investigation on ultrasound images for detection of fetal congenital heart defects

“…Nirmala Devi et al [58] proposed a shift symmetry improved responsive wavelet thresholding feature to reduce speckling in ultrasonic images. In the particular instance of the curvelet shrinkage algorithm, it is a flexible technique that also features an algorithm based on a sub-band adaptive algorithm.…”

Denoising of ultrasound images affected by combined speckle and Gaussian noise

“…The Daubechies complex wavelet transform is used in order to remove speckle noise [14] in which imaginary component of complex scaling coefficient and shrinkage on complex wavelet coefficient are applied, respectively, to detect edges and non‐edges. Improved adaptive wavelet shrinkage is proposed in [15] based on the correlation of the coefficients within and across the resolution scales.…”