Denoising Medical Images using Calculus of Variations

We propose a method for medical image denoising using calculus of variations and local variance estimation by shaped windows. This method reduces any additive noise and preserves small patterns and edges of images. A pyramid structure?texture decomposition of images is used to separate noise and texture components, based on local variance measures. The experimental results show that the proposed method has visual improvement as well as a better signal to noise ratio (SNR), root mean square error (RMSE), and peak signal to noise ratio (PSNR) than common medical image denoising methods. Experimental results in denoising a sample Magnetic Resonance image show that SNR, PSNR, and RMSE have been improved by 19, 9, and 21%, respectively.