基于混合遗传算法的经颅磁刺激线圈阵列优化设计  

Coil Array Optimization Design in Transcranial Magnetic Stimulation Based on Hybrid Genetic Algorithm.

作　　者：杨龙成 陆继庆[1] 张帅[1] 

机构地区：[1]成都信息工程学院,四川成都610225

出　　处：《航天医学与医学工程》2014年第1期32-36,共5页Space Medicine & Medical Engineering

基　　金：国家自然科学基金资助项目(60971043);四川省教育厅重点项目(08ZA098)

摘　　要：目的实现并改善经颅磁刺激(transcranial magnetic stimulation,TMS)的磁聚焦特性。方法构建8个线圈和20个线圈的圆环面线圈阵列模型,结合改进的粒子群算法设计了自适应混合遗传算法,对影响线圈阵列模型磁聚焦性的各种参数进行优化求解。结果在不同平面产生的磁场优化表明了改进的优化算法具有很好的全局搜索与局部搜索能力,在算法的优化下,显示了线圈阵列在优化电流组态下在计算区域内的归一化幅值分布和二维等高线图。结论两种线圈阵列模型在平面上产生了明显的聚焦区域,在目标区域内产生了具有良好磁聚焦性能的磁场分布;在深度方向上有较好的聚焦性能,磁场变化梯度大,更有利于磁刺激兴奋位定点;并且20个子线圈的线圈模型基本上实现了点聚焦,同时具有可聚焦多个靶目标的能力。Objective To achieve and improve magnetic focusing characteristics of transcranial magnetic stimu- lation（TMS）. Methods The toroid coil array models of eight coils and twenty coils were built. Improvement particle swarm algorithm was incorporated into the design of the adaptive hybrid genetic algorithm, and all pa- rameters affecting the coil array model of magnetic focusing were optimized and solved. Results The optimized results of magnetic fields in different plane showed that the improved optimization algorithm was of good search property, and the coil array normalized amplitude distribution and 2D contour plot under the optimized current configuration in computing area with the help of the optimization algorithm. Conclusion Two coil array models could produce obvious magnetic focusing area in the plane and good magnetic field distribution within the tar- get area. It is of better focusing performance in depth direction and magnetic field gradient could change sig- nificantly which is helpful for positioning the magnetic stimulation. The 20 coils model realizes point focusing and could focus on multiole targets at the same time.

关 键 词：经颅磁刺激 磁聚焦 线圈阵列 粒子群算法 混合遗传算法 

分 类 号：R318.6[医药卫生—生物医学工程]