Speckle noise reduction in medical ultrasound images using coefficient of dispersion

Abstract

The major issue that occurs in medical ultrasound (US) imaging is speckle noise, as it degrades the diagnostic quality of images. In this paper, a new method based on the combination of bilateral filter and detail preserving anisotropic diffusion (DPAD) is presented. This combination works on the basis of information obtained as binary classifier map (BCM). A BCM is generated comparing local coefficient of dispersion (CoD) estimated using moving window with the one obtained from a selected homogeneous region. Experiments were conducted on a synthetic kidney US image and a real prostate US image. Results are obtained for proposed method and are compared with the other methods. Objective quality assessment of the proposed method is evaluated on the basis of parameters PSNR, SNR and EKI. The obtained results show the superiority of proposed method over the existing methods.