Medical Image Registration And Fusion

Medical image fusion, a subject which involves information science,computer science and medical imaging, is a new hot spot in medical image processing. As the basic issue of medical image fusion, medical image registration has very important meaning. It is the first step of image fusion, and its precision influenced the effect of image fusion directly.In this paper two rigid registration methods were described. We made researches on the inter-modality medical image registration based on mutual information. Mutual information algorithm use mutual information as a new matching criterion to measure the statistical dependence between the image intensities of corresponding pixels in both images, which is assumed to be maximal if the images are geometrically aligned. The method does not require any prior assumption, medical knowledge, segmentation, feature extraction or other preprocessing steps. Powell algorithm, which does not compute gradients, and Brent algorithm were used to search the optimal rotation and translation parameters. To improve the speed and precision of image registration, we put forward a new method of interpolation, while multi-resolution searching strategy leads to a fast image registration. The registration results of computed tomography(CT) and magnetic resonance(MR) are validated to be sub pixel based on mutual information.After image registration the fusion method is presented. The space-structure fusion and the wavelet transform fusion method are introduced. The space-structure fits a slice of CT image and a slice of MRI image together according to the space. There are no new intensity values which are originally not present in the source image and have two kinds image in the fusion image. The wavelet transformation has several advantages in comparison with the Fourier transformation it can still more and availably describes the image on the hours and the frequency. A series of image, which have different space and distinguish on time and frequency, can be gained by wavelet transformation. These images reflect local changes character of source images, which avail the image fusion. We realize the fusion of CT and MRI based on the wavelets transform.