In Therapeutic images, the analysis activities, for example, object recognition and feature extraction will assume the key part. These errands will end up troublesome if the pictures are defiled with noises. Currently the development of complex algorithms became a new research area. Creating Image denoising calculations is a troublesome errand since fine points of interest in a medical picture implanting analytic data ought not be devastated amid commotion evacuation. A considerable lot of the wavelet based denoising calculations utilize DWT (Discrete Wavelet Transform) in the deterioration stage which is experiencing shift variance and lack of directionality. To overthrow this in this paper we are proposing the denoising strategy which utilizes Undecimated Wavelet Transform to break down the picture and we played out the shrinkage task to wipe out the noise from the picture. In the shrinkage step we utilized semi - soft and stein thresholding functions alongside customary hard and soft thresholding functions and confirmed the reasonableness of various wavelet families for the denoising of therapeutic pictures. The outcomes demonstrated that the denoised picture utilizing SWT (Stationary Wavelet Transform) have a superior harmony amongst smoothness and exactness than the DWT. To survey the nature of denoised pictures, the different measurements we utilized are MSE (Mean Squared Error), SSIM (Structural similarity index measure), UQI (Universal Quality Index), PSNR (Peak signal-to-noise ratio), AD (Average Difference), ST(Structural Content), MD (Maximum Difference). NAD (Normalized Absolute Error).
