Medical Image Registration Algorithm And Application Based On Mutual Information

Image registration is an important technology of image processing technology. In the world, more and more scholars begin to give more attention to the research of image registration. Medical field can produce thousands of medical images because of the rapid development of computer technology and medical imaging technology. Such as CT,X-ray Tomography images, Magnatic Resonance Images(MRI), Position Emission Tomography(PET), Single-Photon Emission Computed Tomography(SPECT), Digital Subtraction Angiography(DSA) and so on. The information is different on relevant organs and tissues of the body because of different medical images. For example, the advantage of CT images is high resolution, can provide anatomical configuration information from internal organs of human; PET and SPECT images can provide functional metabolic information of human body organs and the brain, but low resolution. Therefore, information provided by different images is fused in the field of medical diagnosis. Doctors can obtain more abundant information for clinical diagnosis. For a variety of different, such as image resolution and imaging principle, image registration is the premise of image fusion, keeping the two points from the two images is consistent in Geometry location. So it is necessary to study image registration technology.How to measure the effects of Image registration, as one of the most commonly used in image registration evaluation standards, mutual information don’t need to image feature point extraction and re-processing. It has been appreciated by researchers as its advantages, such as less manual Intervention, good robustness, high degree of automation. And the registration accuracy is higher than the method based on other criteria. In this paper, the image registration algorithm which is based on mutual information is studied and applied. The main work is as follows.(l)It detailed introduces and implement the principles and methods which use mean filtering, median filtering and wavelet transform in medical image denoising. The common geometric transformation between two images are examples.(2)First, It detailed introduces the principle of image registration. POWELL algorithm and PSO algorithm about image registration algorithms which are based on mutual information are studied. Second, a kind of image registration method based on wavelet transform is researched, which realize Image registration by wavelet transform of image decomposition. Finally, our goal is high registration precision, fast speed and strong robustness by comparison and analysis of the results of various algorithms.