A Review on CT and X-Ray Images Denoising Methods

Abstract

In medical imaging systems, denoising is one of the important image processing tasks. Automatic noise removal will improve the quality of diagnosis and requires careful treatment of obtained imagery. Computed tomography (CT) and X-Ray imaging systems use the X radiation to capture images and they are usually corrupted by noise following a Poisson distribution. Due to the importance of Poisson noise removal in medical imaging, there are many state-of-the-art methods that have been studied in the image processing literature. These include methods that are based on total variation (TV) regularization, wavelets, principal component analysis, machine learning etc. In this work, we will provide a review of the following important Poisson removal methods: the method based on the modified TV model, the adaptive TV method, the adaptive non-local total variation method, the method based on the higher-order natural image prior model, the Poisson reducing bilateral filter, the PURE-LET method, and the variance stabilizing transform-based methods. Our task focuses on methodology overview, accuracy, execution time and their advantage/disadvantage assessments. The goal of this paper is to provide an apt choice of denoising method that suits to CT and X-ray images. The integration of several high-quality denoising methods in image processing software for medical imaging systems will be always excellent option and help further image analysis for computer-aided diagnosis. Povzetek: Pregledni članek opisuje metode za čiščenje slike, narejene z rentgenom ali CT.

Enhancing industrial X-ray tomography by data-centric statistical methods

“…In X-ray tomography, the noise process e is often modeled as a compound Poisson distributed noise (Whiting, 2002;Thanh et al, 2019). We make a simplification and assume that e is zero-mean Gaussian white noise with covariance C ¼ σ 2 I.…”

Diagnosis of breast cancer based on modern mammography using hybrid transfer learning

“…Mammography, a special X-ray of the woman’s breast, is one of the most common diagnostic tools for detecting breast cancer (Bharati et al 2020a , b ; Thanh and Surya 2019 ). A 3D mammography is an advanced model compared to mammography.…”

Adaptive frequency median filter for the salt and pepper denoising problem

“…There are many types of noise such as Gaussian noise [10], Poisson noise [11–13] and impulse noise [14–18]. Salt‐and‐pepper noise (SPN) is a simple form of impulse noise and it is common in many electronic devices.…”