Different denoising techniques for Medical images in wavelet domain

Diagnosis of Medical images is very difficult when images are corrupted with noises during the process of acquisition. Now a days development of effective algorithms for removal of noise has become an important research area. Developing Image denoising algorithm is a difficult task since fine details in a medical image embedded with diagnostic information should not be destroyed during noise removal. Most of the existing denoising algorithms use DWT but it has the drawback of shift variance. To overcome this, here the denoising method which uses Undecimated Wavelet Transform to decompose the image has been proposed and the shrinkage operation such as semi-soft and garrote thresholding operators along with traditional hard and soft thresholding operators are used. The suitability of different wavelets for the de-noising of medical images using performance indices SSIM, PSNR and MSE are tested.