A comparative analysis of preprocessing techniques on ultrasound images of CCA

Stroke stands as a leading contributor to global mortality, with a substantial rise in human deaths attributable to cardiovascular diseases. Ultrasound imaging serves as a valuable tool for atherosclerotic plaque diagnosis, as plaque accumulation plays a pivotal role in the development of cardiovascular diseases, which can ultimately lead to fatal outcomes. Within this framework, a systematic and rational approach is essential for the identification and diagnosis of carotid plaque. This logical process will help to analyze and identify unrevealed data hidden in the ultrasound images of the common carotid artery. Several methods are applied to detect the presence of plaque within the common carotid artery. The primary goal of this paper is to give a widespread review of the filtering techniques and methods to reduce or eliminate the speckle noise to a certain extent in the ultrasound images of the common carotid artery. Although ultrasound imaging is one of the non-invasive and economical techniques, the obtained ultrasound images have low quality in terms of contrast, resolution, and sensitivity. To recover high quality images from the noise images, preprocessing are performed for de-noising the images in addition to image enhancement and restoration, which helps improve the ultrasound image quality. The comparative analysis of various preprocessing methods involves the assessment of performance indicators such as Peak Signal to Noise Ratio, Signal to Noise Ratio, Speckle Suppression Index, and Mean Square Error. Ultrasound B-mode carotid artery images underwent noise reduction using a range of techniques, including Average, Median, Gaussian, Anisotropic, Bilateral, Wiener, Lee, Wavelet, Total Variation, and Block matching 3D filtering methods. The goal was to diminish noise while retaining essential image features, a challenge addressed by various strategies within the existing literature. Each method carries its advantages and drawbacks. In this article, we detailed several significant studies in the realm of image de-noising. Initially, we introduced the problem of image de-noising and then proceeded to outline various de-noising techniques. Our discussion included the distinctive attributes of these techniques. To assess the effectiveness of different preprocessing strategies, we utilized performance metrics such as Mean Square Error (MSE), Peak Signal to Noise Ratio (PSNR), Signal to Noise Ratio (SNR), and Speckle Suppression Index (SSI). Ultimately, we conducted a comparative analysis of the performance of traditional de-noising filters and presented the results.