Medical Image Registration Based On Improved Demons And NCQPSO Algorithm

Medical image registration is to find a geometric transformation that maps a given moving image to a fixed image to make them most similar. The paper firstly introduces current condition of medical image registration, and analyzes the procedure of medical image registration. In addition, the paper analyzes current primary medical image registration algorithms, and reviews study of the current elastic registration status at home and abroad.Due to mutual information measure for the large amount of computation, long registration time, more local extremum problem, firstly medical image registration method Buzug proposed was improved. The new measures were derived by replacing the Shannon entropy function in mutual information with any strictly concave function, which were named mutual strictly concave function measures. According to Jensen inequality and Schur concave function, a new measure of image registration were proposed-Jensen-Schur measures (JS), which enlarged the definition of mutual strictly concave function measures and whose independent variable was a vector. And then proved new constructions of JS2, JS3, JSω measurement are better than mutual information measure(MI), normalized mutual information measure (NMI) in the calculation speed and convergence performance and noise immunity through experimental verification. Finally, the JS2measure for multi-modal medical image registration experiments has good registration results.This paper uses Niche chaotic mutation quantum-behaved particle optimization (NCQPSO) as medical image registration algorithm for optimal search algorithm. In this algorithm, niche methods and eliminating strategy are introduced to improve the global optimizing ability. Futher, shrinking chaotic mutation, which behaves well in refined local traversal searching, is introduced to improve the precision. Simulations show that NCQPSO algorithm presented can avoid the premature convergence of standard PSO algorithm effectively and has powerful optimizing ability, good stability and higher searching precision. It is also superior to traditional QPSO.Based on the successful application of JS measure in multi-modality image registration, an improved Demons algorithm for elastic multi-modality images is proposed in this paper. The method adds additional external force defined as the gradient of JS measure between two images with respect to the deformation fields to drive the floating image to deform. In this way, the misregistration problem resulted by the original algorithm when transformation direction can not be determined due to the lack of intensity gradient information can be overcome. According to the improved Demons image registration algorithm combined with NCQPSO optimization algorithm, the method has good registration results in medical image registration process. Experiment also compares the PSO, QPSO and NCQPSO registration results. The experimental results show that the NCQPSO algorithm as optimization strategy, can solve the registration problem in global optimization with good accuracy and robustness. A variety of experiments show the proposed method is an effective, fast and accurate registration.