Medical Image Registration Method

With the rapid development of medical information technology, computer image technology has been widely applied in medical image processing, image storage and management. Medical image registration is an indispensable step in surgery and subsequent treatment of the track, it is not only corrected registration the image, increase the amount of data in real-time during the procedure, and enhance the success rate of surgery greatly, protect the life and health of patients, but also allow medical staff to avoid being intrusive rays. Registration of the two medical images with different sizes and characteristics has become an urgent problem in current medical image research.Image registration is an important part of medical image research. Medical image registration can be divided into rigid image registration and non-rigid image registration based on the desired image registration. According to different subjects, the registration can be divided into two forms: registration between same individuals and registration between different individuals. The registration algorithm mainly can be divided into three categories: image registration based on processing flow, image registration based on image characteristics and image registration based on deformation model. Compared with rigid registration, non-rigid registration in clinical medicine application is broader, technology is more difficult, requirement is higher, it also has more important application significance.The main research work of this paper is as follows:1. We have made a research for the principles of B-spline function. By analyzing the deficiency of registration accuracy with B-spline function in non-rigid registration and related problems, we introduced a image processing method of with NSST. With the proposed algorithm, images are transformed by NSST, the registration is implemented by using different registration methods for the high-frequency and low frequency of transform image respectively. Experiment results show that the proposed method outperform the method with B-spline. The proposed algorithm not only has better robustness, higher registration accuracy, but also has higher computational efficiency. 2. We have made detailed research of recent related issues of medical image registration techniques. The analysis for development, the difficulties and the focus of the registration work in the future of home and abroad are presented with the research on the three aspects which include the classification of registration, the development of a specific model and the application of registration methods. Research in medical image registration is summarized, and various elastic models are studied and summarized respectively.