Research On Image Fusion Technology Of CT And MRI

With the development of the computer technology, medical imaging technology, information technology, medical image fusion technology becomes a research hotspot. Make use of medical image fusion technology process redundant data and complementary information in multiple images to improve the reliability and the clarity of the image.Firstly, the paper introduces the background of the medical image fusion technology, and elaborates the basic knowledge, basic treatment method, registration, fusion algorithm involved in the field of medical image fusion. In terms of image preprocessing, the paper researches on frequency domain enhancement and spatial domain enhancement, through MATLAB simulation test and verify various images processing algorithm, summarize the advantage and disadvantage of different edge detection operator.Secondly, in terms of image registration, the paper elaborates medical image registration's basic concepts and registration's evaluation; analyzes the theory of registration; in-depth studies of the common method of medical image registration; proposes the image registration optimization algorithm based on maximum mutual information. According to the characteristics of wavelet sampling, make use of this algorithm execute wavelet stretching, rotation and translation transformation. In the search for optimization, the paper adopts Powell direct search method, interpolation method adopts PV interpolation. Through MATLAB simulation, wavelet maximum mutual information registration algorithm reduces the registration time and it is able to achieve sub-pixel registration.Finally, in terms of image fusion, the paper streamlines the original PCNN model, form simplified PCNN model, and combines it with the wavelet algorithm. Through study and compare various fusion methods, prove wavelet PCNN algorithm is more superior to other fusion algorithms.