Annu Int Conf IEEE Eng Med Biol Soc. 2021 Nov;2021:4337-4340. doi: 10.1109/EMBC46164.2021.9629971.
Transcranial direct current stimulation (tDCS) delivers weak current into the brain to modulate neural activities. Many methods have been proposed to determine electrode positions and stimulation intensities. Due to the trade-off between intensity and focality, it is actually a multi-objective optimization problem that has a set of optimal solutions. However, traditional methods can produce only one solution at each time, and many parameters need to be determined by experience. In this study, we proposed the nondominated sorting genetic algorithm II (NSGA-II) to solve the current optimization problem of multi-electrode tDCS. We also compared the representative solutions with LCMV solutions. The result shows that a group of solutions close to the optimal front can be obtained just in only one run without any prior knowledge.
经颅直流电刺激（tDCS）将弱电流传入大脑以调节神经活动。已经提出了许多方法来确定电极位置和刺激强度。由于强度和聚焦之间的权衡，这实际上是一个具有一组最优解的多目标优化问题。然而，传统方法每次只能产生一个解，并且许多参数需要凭经验确定。在这项研究中，我们提出了非支配排序遗传算法 II（NSGA-II）来解决多电极 tDCS 的电流优化问题。我们还将代表性解与 LCMV 解进行了比较。结果表明，仅在一次运行中就无需任何先验知识即可获得一组接近最优前沿的解。
