Speckle Noise Filtering for Ultrasound Images of Common Carotid Artery:A Review

Speckle is modeled as a signal dependent noise, which tends to reduce the image resolution and contrast, thereby reducing the diagnostic values of the ultrasound imaging modality. Reduction of speckle noise is one of the most important processes to increase the quality of biomedical images. Filters are used to improve the quality of ultrasound images by removing the noise. This paper compares the performance of the thresholding technique Bayes Shrink in despeckling the medical ultrasound images with other classical speckle reduction filters like Lee, Frost, Median, Kaun, Wavelet Bayes, Anisotropic diffusion and Wavelet. The performance of these filters is analyzed by the statistical measures such as Peak Signal-to Noise Ratio, Mean Square Error and Equivalent Number of Looks. To produce a better quality resolution picture, the filter should have high Peak Signal to Noise Ratio, low Mean Square Error, high Equivalent Number of Looks. The results obtained are presented in the form of filtered images, statistical tables and graphs. Finally, the best filter has been recommended based on the statistical and experimental results. From the results obtained Lee and Frost filter outperforms the other mentioned filters in terms of high PSNR and low MSE for high variance of noise where as anisotropic diffusion filter outperforms with high PSNR and low MSE with maximum ENL for low variance values of noise.