基于互信息的医学图像配准中互信息的计算_张见威精选.pdfVIP

25 1 Vol. 25 No. 1 2008 2 Journal of Biomedical Engineering Februar 2008 * 张见威 韩国强 ( 华南理工大学 计算机科学与工程学院, 广州51064 1) , , , , , , MRI CT , 32 64 , 256 , , TP391; R31804 A 1001-5515(2008) 01-0012-06 Computation of Mutual Information in Medical Image Registration Based on Mutual Information Zhang Jianwei Han Guoqiang ( School of Computer Science and Engineer ing, South China University of Technology , Guangzhou 510640, China) Abstract Registration ased on mutual information is a t ypical method in medical image registration. Mutual infor- mation is a common similarity measure in image registration, which has excellent ro ustness and accuracy , ut large calcu- lation amount makes it difficult to e applied to clinics. In this paper, a registration algorithm ased on mult-i resolution and hy rid optimization is adopted to implement 2-dimension monomodal and multimodal registrations of MRI and CT im- ages of human heads w ith different num ers of gray ins. Results of experiments show that registration precisions have not nota le change with 32, 64 gray ins, compared with 256 gray ins, w hereas the computation costs decrease remarka ly. Ke words Medical image registration Mutual information( MI) Num er of gray ins Mult-i Resolution Hy rid optimizat ion : 1 , , , , , , , , [ 1~ 3] [4~ 6] , ( ) ; : ( Simulated Annealing , Algorithm, SAA) ( Part-i , cle Swarm Optimization, PSO) Pow ell , , , ; , , , , , , ,