Performance evaluation of filters for de-noising the intravascular ultrasound (IVUS) images

Abstract

Intravascular ultrasound is a catheter based ultrasound imaging used in the diagnosis and treatment of cardiovascular disease. It facilitates high-resolution tomography assessments of the dimensions of vessels, as well as direct examinations of the vessel walls and underlying atherosclerotic plaque morphology. Ultrasound imaging is degraded by the presence of signal-dependent noise and artifacts that lead to incorrect image measurements. The purpose of de-noise is to remove the speckle noise and the artifacts as well as retaining the edges and other detailed features as much as possible. In this paper, the performance of different filtering techniques such as median, anisotropic, wavelet, and homomorphic are analyzed and evaluated with quantitative measurements viz. mean square error (MSE), peak signal to noise ratio (PSNR) and signal to noise ratio (SNR).