基于模拟退火算法的磁感应治疗热籽分布

Distribution of thermo seeds in magnetic induction therapy based on the simulated annealing algorithm

肿瘤区域内的热籽适形分布算法是肿瘤磁感应治疗计划系统中的一个关键技术,其核心问题是使用最小数量的热籽达到预期的温度场分布。为解决热籽适形分布问题,该文提出了结合热籽产热传热学的模拟退火算法。分析了应用模拟退火算法解决热籽分布问题的计算方法,包括热籽分布模型的设计、产热传热模型的建立以及退火计划的设计。考虑到热籽分布问题的特殊性,重新定义了初始温度与停止条件、评价函数与接受概率、退火策略以及具体计算步骤。通过计算机仿真对肿瘤区域建模与网格化、热籽植入模型、温度场分布以及算法迭代进行了模拟,验证了应用模拟退火算法解决热籽分布问题的可行性。

Optimum distribution of the thermo seeds in a tumor is key to magnetic induction therapy planning. The core objective is to achieve the desired temperature distribution using the minimum number of thermo seeds. The optimum distribution of the thermo seeds is found using the simulated annealing （SA） algorithm combined with bio-heat production and transfer theory. This paper describes the calculational method to solve for the distribution of thermo seeds using the SA algorithm, including the design of the thermo seed distribution model, the bio-heat production and transfer model and the SA scheme. The algorithm is designed for the thermo seed distribution, initial temperature, stop condition, evaluation functions, acceptance probability and annealing strategy. The modeling and meshing of the tumor region, the modeling of the thermo seed implantation, the temperature distribution, and the iteration process are all introduced to show the feasibility of solving for the thermo seed distribution using the SA algorithm.