Medical image denoising using multi-resolution transforms

•Wavelet, Curvelet, Ridgelet, Contourlet based multi-resolution transforms which is used for denoising the medical image.•Applied on the Gaussian, Rician and Rayleigh noises added image.•GLRLM is the quantity of keeps running with pixels of gray level ‘i’ and run length ‘j’ for a provided guidance. Medical Imaging techniques are most commonly used by radiologists for visualizing detailed internal structure and function of the body. During such imaging process the techniques often have a complex type of noise due to patient’s movement, transmission and storage devices, processing and reconstruction algorithms. Three types of noises considered in this paper. They are Gaussian noise, Rician noise and Rayleigh noise which are added to the medical image. Then the different image transformation techniques like Wavelet, Curvelet, Ridgelet and Contourlet Transform etc., can be used for denoising purposes with each of the transform techniques having its own significance. From the denoised image of each transforms, the features are extracted from the image using feature extraction techniques. Three types of feature extraction methods such as Gray Level Run Length Matrix (GLRLM), Gray Level Co-occurrence Matrix (GLCM) and Markov random field are used to extract the features of the image. Performance is analyzed based on the values of Mean Square Error, Signal to Noise Ratio, SSI, PSNR and Visual Evaluation. In this paper the ridgelet transform provide better estimate of MSE, SNR, SSI value for Gaussian (37.56, 5.95, 0.9), Rician (32.68, 16.55, 0.92) and Rayleigh noise (260, 7.54, 0.88).