Research Of Non-Rigid Registration Method Based On 2D-3D Images

In traditional surgery, the doctor analyze the illness according to some modal image, which is a part of patients sits, such images include CT images, MRI images and X-ray images. More often, in order to improve the level of diagnosis, two or more than two kinds of different modal images are combined to analyze judge. In this process, the doctor can locate a part of the patient more accurately by using the registration between two-dimensional X-ray images and three-dimensional CT data, which is a very important step to minimally invasive surgery. The main research work of this paper is 2D-3D image registration, the research work mentioned in the paper is mainly about:(1) Introduces the research status and problems of registration technology at home and abroad, and expounds the basic principles and methods of medical image registration.(2) The image noise and three kinds of denoising technology are introduced, this paper improves the quality of the image by filtering denosing on original image, and has laid a good foundation for the later image registration.(3) Discusses digitally reconstructed radiographs, we obtain 2D projection image by using the ray casting algorithm to reconstruct the 3D CT volume data, which transforms 2D-3D registration into 2D-2D registration. This method reduces the complexity of registration, and improves the registration efficiency.(4) The experimental material, hardware and software environment, and experimental methods in this paper are stated, in particular, the basic principles and steps of 2D-3D image registration are expounded emphatically. This paper researches a registration method based on feature, and detects corner points by using Harris operator. In addition, a relative position function are proposed in this paper, which detects corner points to match precisely, and the relevant experiments are carried out to analyze and discuss the proposed method.From the experimental results we can see that the proposed method has better treatment effect, high operation, high precision and good comprehensive performance.