A Brief Review: Speckle Filtering Methods

Abstract

Due to non-invasive nature, low cost, capability of real time imaging formation and the on-going improvements in image quality of Ultrasound (US), it is a widely used and safe medical diagnostic technique. Irrespective of these features of US imagery, its main disadvantage is the presence of noise due to backscattered echo signals called speckle. Therefore, it is not only difficult for the physician to analyze and diagnose the image in the presence of speckle noise but feature extraction, and further analysis is also difficult. There are numerous techniques found in the literature to reduce the speckle noise in US images, it is hard to find out the best techniques among them and to get the drawbacks of existed methods with a comparison to each other. Generally there are two types of techniques (spatial and frequency domain techniques) depending upon the domain used while calculating the modified pixels for speckle reduction. In this paper, performance of various spatial and frequency domain speckle reduction techniques is compared  using quality metrics (EPF, SSIM, SNR, and RMSE) and it is concluded that SRAD filter outperforms all other speckle reduction filters