Kalman filter based estimation of ionic concentrations and gating variables in a cardiac myocyte model

The purpose of this study is to give researchers improved access to important electrophysiological quantities, such as ion-channel gating variables, that are difficult or impossible to measure during in vitro experiments, yet are thought to be critical to the formation of dangerous arrhythmias. To help fulfill this goal, we examined the feasibility of inferring these types of quantities from more readily available data, such as measurements of cellular membrane potential. First, we performed an observability analysis on a linearized Luo-Rudy dynamic (LRd) myocyte model, which showed that concentration and gating variables in the LRd model can be reconstructed from membrane potential data. Next, we designed a Kalman filter for the model and tested its performance under simulated conditions. The tests demonstrated the ability of the filter to produce more accurate estimates of the system state compared to the case without measurement feedback. This research is relevant to human health, since state estimation methods such as Kalman filtering could be used to obtain more information about the response of a single cell to the influence of pharmacological agents or other antiarrhythmic therapies, over a larger range of cellular variables than are typically monitored during an in vitro study.