Medical Image De-Noising Using Combined Bayes Shrink and Total Variation Techniques

Digital image processing brought a tremendous revolution in many fields; medical field being one of them. Medical images are used to detect abnormalities and diseases in the human body simply by analyzing them. During the acquisition process, noise signals may be introduced in these images which will impact negatively on the diagnostic process. Noise signals degrade the image quality by suppressing useful information present in the form of edges, fine structures, textures and so on. Hence it is essential to suppress noise signals because noisy images may lead to false interpretations by the radiologists. Suppression of noise signals from the medical images is called "Medical Image De-noising". Some examples of medical images are Magnetic Resonance Images (MRI), Computed Tomography (CT) images, Ultra-Sound (US) images and so on. These images are corrupted by various noise signals; for example, MRI images are affected severely by noise which is known as Rician noise. CT images are corrupted by Gaussian noise, while US images are affected by multiplicative noise called speckle noise. To remove these unwanted noise signals various filters have been proposed, but none of these methods can be used as a global de-noising technique because any filter which can remove one noise effectively fails to remove others. Hence, it is necessary to develop a filter that can remove many noise signals because any image may be corrupted by more than one noise. This requirement motivates the researchers to push ahead to achieve such goal. In this chapter; a framework has been proposed to de-noise medical images, reducing the effect of additive white Gaussian noise. It consists of various spatial domain filters, specifically median filter and median modified Wiener filter. It also uses adaptive wavelet thresholding and total variation technique in parallel, the results of which are then fused together using wavelet based fusion technique. This process is named as shrinkage combined enhanced total variation technique as it enhances the quality of de-noised images. Digital image processing brought a tremendous revolution in many fields; medical field being one of them. Medical images are used to detect abnormalities and diseases in the human body simply by analyzing them. This chapter introduces a framework to de-noise medical images, reducing the effect of additive white Gaussian noise. It consists of various spatial domain filters, specifically median filter and median modified W