A New Transform for Medical Image Denoising: Fused Tetrolet Transform

Denoising is a very important task of medical image processing. Noiseless images provide more accurate diagnosis. Using multi-resolution analysis methods for carrying out medical denoising has become more popular in recent days. Tetrolet transform which is an adaptive form of Haar wavelet transform is one of these methods. In this study, a new form of Tetrolet called "fused Tetrolet transform" is proposed. Denoising performance of this novel method is compared to Wavelet transform, standard Tetrolet transform and four other types of Tetrolet transform. Fused Tetrolet and the modified Tetrolet types are initially used for denoising. In this study, random, gaussian and poisson noises are added on 30 liver magnetic resonance (MR) images and 30 mammography images separately and then removed. Peak signal to noise ratio and structural similarity index are used as evaluation criteria. The results show that fused Tetrolet transform surpasses the four modified forms of Tetrolet, standard Tetrolet and Wavelet while denoising both mammography and liver MR images for all types of noise.