基于随机漂移粒子群优化算法的三维脑部磁共振图像分割  

3D MR Brain Image Segmentation Through Random Drift Particle Swarm Optimization-Based Algorithm

作　　者：施佳佳[1] 孙俊[1] 范方云[1] 王梦梅[1] 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122

出　　处：《江南大学学报（自然科学版）》2015年第4期403-407,共5页Joural of Jiangnan University (Natural Science Edition)　

基　　金：国家自然科学基金项目(61170119)

摘　　要：设计并验证了一种应用于马尔科夫随机场的脑部磁共振图像的分类模型。该模型是基于随机漂移粒子群优化算法。它相对于标准粒子群算法在收敛速度和优化性能方面都有显著提升。针对模拟和临床脑部磁共振图像的实验结果表明,该模型与模拟退火、标准粒子群算法相比,具有更好的收敛性能和分割精度,是一种有效的图像分类模型。Markov Random Field( MRF) model is a classical image analysis method in computer vision,which has been widely utilized in image segmentation. This paper designs and validates a novel segmentation method for the magnetic resonance image of brain tissues based on MRF. The method is inspired by the previously developed random drift particle swarm optimization( RDPSO) algorithm,comeing from the study of the free electron model. The RDPSO is a stochastic optimization algorithm,which shows better performance than the standard particle swarm optimization algorithm. Compared with the simulated annealing algorithm and particle swarm optimization algorithm,the experimental results of the simulated and real MR images show that the proposed method is effective for image classification models,which possess better convergence ability and classification precision.

关 键 词：马尔科夫随机场 磁共振图像分割 随机漂移粒子群优化算法 分类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]