用于伽玛刀手术设计的算法研究  

Algorithms for design of Gamma Knife

作　　者：王钦平[1] 季春霖[1] 孙雷[1] 刘思齐[1] 郝培锋[1] 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110004

出　　处：《哈尔滨工业大学学报》2003年第11期1317-1319,共3页Journal of Harbin Institute of Technology

基　　金：国家自然科学基金资助项目(69974008);21世纪初中国高等教育人才培养体系研究计划资助项目(C369).

摘　　要：提出一种基于腐蚀算法和遗传算法的伽玛刀手术设计方法.首先用腐蚀算法,找到靶点中心的可行区域,将靶点限制在目标体之内,然后用遗传算法寻找每个靶点的半径和靶点中心的最佳位置.在遗传算法中,重点研究了基因编码,初始种群,适应值函数和交叉操作.该算法不要求被寻优函数连续可微,其增加的计算量不大.使用该算法进行的手术设计算例表明,靶点能够覆盖目标体体积的90%以上,能够满足实际手术的要求.The possible location of shots' centers is obtained by using the Erosion method, and the shots are then constrained within the area. The Genetic algorithm is used to search for the possible location of shots' centers, and find out the radius and the best location of every shot. The Genetic Algorithm is studied with emphasis on encoding scheme, initialization, evaluation function and crossover operation. The operator does not require that the function to be optimized continuous or differentiable, and the increased computation demand is little. Some examples designed using the approach are given to show that the shots can cover more than 90% of the target volume.

关 键 词：伽玛刀手术设计 腐蚀算法 遗传算法 基因 脑部肿瘤 放射治疗装置 靶点 

分 类 号：R730.55[医药卫生—肿瘤]