This work presents new adaptive time-step schemes based on the Uniformization method for solving stiff electrophysiology models of excitable cells. Uniformization is a numerical method for continuous-time Markov chains that was proven to be effective in solving modern membrane models that have ion channels described by Markov chains. The new methods were tested using three highly stiff cell models previously published. In addition to the simulation of single cells, simulations of well connected cellular networks were performed using the monodomain model. Our results suggest that the proposed adaptive numerical methods significantly outperform previous techniques and are up to 27 times faster than the classical adaptive time-step Euler method.

这项工作提出了基于均匀化方法的新的自适应时步方案，用于解决可兴奋细胞的僵硬电生理模型。均匀化是连续时间马尔可夫链的一种数值方法，事实证明，该方法可有效解决具有马尔可夫链描述的离子通道的现代膜模型。使用先前发布的三个高刚性单元模型测试了新方法。除了模拟单细胞外，还使用单域模型对连接良好的蜂窝网络进行了模拟。我们的结果表明，提出的自适应数值方法明显优于以前的技术，并且比经典的自适应时步欧拉方法快27倍。