Parameter Estimation of Ventricular Myocardial Cell Model Using an On-Line Learning Algorithm

The Luo-Rudy dynamic (LRd) model is the one of typical models of ventricular myocardial cell and described by Hodgkin-Huxley-type nonlinear ordinary differential equations. By changing various parameters of the LRd model, we can reproduce heart conditions, which trigger heart diseases such as arrhythmia. It is, however, very difficult to understand the relation between the parameters in the LRd model and a heart cell´s behavior (action potential) because the LRd model has high complexity and nonlinearity. We demonstrate to estimate the parameters in the LRd model easily and automatically with a learning algorithm. Thus we show that the automatic parameter estimation is very useful to identify the cause of heart diseases in clinical applications