Noise suppression using Wiener filtering in the nonsubsampled contourlet domain for Magnetic Resonance brain images

Abstract

The Magnetic Resonance Imaging (MRI) is becoming famous in medical imaging because of its diagnostic applications in the medical analysis and benefits over other medical diagnostic methods. However, it is observed that the diagnosis operations is becoming difficult when the noise gets introduced in MR brain images as noise is the significant factor which influences the quality of diagnosis and treatment process of brain tumors and hence noise minimization is of significant concern. The predominant noise in MRI is Rician noise, which is the signal dependent noise while in this paper the Rician noise minimization is attained by combining the Wiener filter and Non Subsampled Contourlet Transform(NSCT), which conserves excellent characteristics of the MR brain image. For measuring the performance of proposed method, the Peak Signal to Noise Ratio (PSNR), Coefficient of Correlation (CoC) and Quality Index (QI) are utilized in comparison with denoising by Wiener filter in Wavelet domain and Contourlet domain. It is evaluated that the proposed method has provided good results in the chosen evaluation metrics.