3D CT Denoising by New Combination Between Nl-Mean Filter and Diffusion Tensor

The images acquired from different techniques of medical equipment are generally noisy data. The noise distribution in the CT image is modeled as a Gaussian distribution which appears in the images as a random fluctuation allowing to a misdiagnosis. So the denoising of the CT images is a challenge task in medical area. In this paper we propose a new denoised method based on combination between the Non local mean filter and the Diffusion Tensor for 3D Computed tomography scan data with a MAD estimator for gaussian noise. A quantitative measures was calculated and compared to other common denoising methods, improving the efficiently of our algorithm in term of removing noise and preserving significant details.