The Pde-based Medical Image Enhancement Technology

With medical imaging technology being used in the medical diagnostic process, accuracy and timeliness of medical diagnostic has been greatly improved.Medical images is playing an increasingly important role and has an irreplaceable position in clinical diagnosis. Thus, a unique image processing (Medical Image Processing) arose. It is an important branch of digital image processing. And It's main job is to process and handle a variety of medical images. Processed images more easily used in clinical. There are X-ray maging (X-CT), magnetic resonance imaging (MRI) and nuclear medicine imaging (NMI) and so on. Due to some unfavorable factors such as the imaging system and imaging equipment and the body's own organization, the image we get is very different from the one we expected. Degraded image is mainly manifested as containing noise, the details blur, contrast and poor. And it will significantly impact the clinical analysis and diagnosis of the doctor, and also even result in misdiagnosis. Therefore it is necessary for us to filter out the noise in medical images and enhance image contrast by studying effective pretreatment method. This article focuses on the medical image enhancement processing technology based on partial differential equations.In this paper, the two researches of medical images based on partial differential equations are described, one of which is anisotropic diffusion model and the other is the total variation denoising model. The paper analyzes the advantages and disadvantages, diffusion properties of the models. As both the anisotropic diffusion and total variation denoising models can only denoising and edge preserving, a new adaptive variational denoising model for medical image enhancement is presented by employing local times gradient fidelity term, and another new medical image enhancement based on partial differential equations is proposed by coupling shock filter. The new models not only can effectively filter out noise, but also enhance high-frequency region of medical image such as the edges and details. In a certain extent, the new models also can overcome the negative edge preserving problems of original models and improve the visual effect of medical images.In addition, medical images in actual clinical often contain noise, which, inevitably led to details of the fuzzy and the low contrast of the medical images. Because most of denoising enhance models are just effective to enhanceing the medical image edges and details, and are not ideal to enhancement of the texture, in this paper, a new synchronization method for combining denoising with contrast enhanced based on partial differential equations is proposed. The new method make the process of denoising and enhancement be disposed simultaneously to the medical image by coupling diffusion operator and the classical histogram equalization method. By this means, the low contrast of interior and edge of the area which we are interested in is ameliorated, and image information becomes richer. The new method also add a special adjustment item for eliminating ladder effect to further improve the image visual effect.