Denoising of MRI Images Using Curvelet Transform

Abstract

Most of the medical images are usually affected by different types of noises during acquisition, storage, and transmission. These images need to be free from noise for better diagnosis, decision, and results. Thus, denoising technique plays an important role in medical image analysis. This paper presents a method of noise removal for brain magnetic resonance imaging (MRI) image using curvelet transform thresholding technique combined with the Wiener filter and compares the result with the curvelet and wavelet-based denoising techniques. To assess the quality of denoised image, the values of peak signal-to-noise ratio (PSNR), mean square error (MSE), and structural similarity index measure (SSIM) are considered. The experimental results show that curvelet denoising method depicts better result than wavelet denoising method, but the combined method of curvelet with Wiener filtering technique is more effective than the wavelet- and curvelet-based denoising method in terms of PSNR, MSE, and SSIM.