基于生物地理学优化算法的医学图像配准  

Medical image registration based on biogeographical optimization algorithm

作　　者：周全 于志强 ZHOU Quan;YU Zhi-qiang(Information and Network Center,Yunnan Minzu University,Kunming 650500,China)

机构地区：[1]云南民族大学信息与网络中心,云南昆明650500

出　　处：《云南民族大学学报（自然科学版）》2020年第6期614-618,共5页Journal of Yunnan Minzu University:Natural Sciences Edition

基　　金：云南省教育厅科学研究基金(2020J0360)。

摘　　要：基于信息强度的医学图像配准主要有相似性测度、插值计算、空间变换以及寻优算法几个主要步骤组成.在完成医学图像配准的过程中,性能优异的寻优算法可以极大程度提高医学图像配准的配准精度和配准时间.使用了生物地理学优化算法作为优化算法,粒子群算法作为对照优化算法,在阿尔茨海默病成像计划数据集进行了一系列的单模态刚体图像配准实验.实验结果表明,算法可以有效地提升医学图像配准的精度与速度,鲁棒性也更强.Medical image registration based on information intensity mainly consists of similarity measure, interpolation calculation, space transformation and optimization algorithm. In the process of medical image registration, the high-performance optimization algorithm can greatly improve the registration precision and registration time. In this paper, a series of unimodal rigid-body image registration experiments are carried out by using the biogeographical optimization algorithm and the particle swarm optimization algorithm as the control optimization algorithm in the imaging plan dataset of Alzheimer’s disease. Experimental results show that this algorithm can effectively improve the accuracy and speed of medical image registration, and the robustness is also stronger.

关 键 词：生物地理学 优化算法 医学图像配准 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术]