Research Of Magnetic Resonance Images Registration

Medical image registration and fusion is a crossing research topic of information science, computer image technology and modern medicine. Retrospective medical image registration is a key technology to make full use of the modern multimodality medical images integratedly. It is very important to the automatic analysis of serial images and multimodality images, and widely used in the areas of clinical diagnoses, therapy, quality assurance and evaluation.Firstly, related concept, methods and the up-to-the-date situation are briefly reviewed in this thesis. Then several commonly-used similarity measures were compared under multiresolution wavelet framework. A new registration method was released, which is using different similarity measure on different level of wavelet pyramid. The experimental results show that the proposed method obtained a longer capture range and more accurate results.Secondly, we proposed an image registration approach based on curvelet transform. Compare with the classical wavelet transform, curvelet transform is more appropriate for the curve analyzing and noise reduction. Experimental results show that the proposed algorithm can achieve subpixel precision with higher accuracy and robustness than wavelet transform based algorithm.Finally, this thesis proposed an improved PSO-GA optimum. PSO-GA hybrids offer a combination of the GA (Genetic Algorithms) and PSO (Particle Swarm Optimization) with the hope of utilizing the GA operators. Experimental results show that the PSO-GA optimum can avoid the local maxima of similarity measures effectively, and improved the accuracy of registration process.