In silico electrophysiological models are generally validated by comparing simulated results with experimental data. Markers for validation include action potential duration (APD), maximum transmembrane potential (Vmax), triangulation and others. The reference values for some markers are obtained from tissue preparations, and their values may be strongly influenced by inter-cellular conduction. Not accounting for conduction effects may introduce a source of error when comparing simulations with experimental observations, since for some markers there is no close correspondence between cell and tissue values. Our hypothesis is that the cell-tissue mismatch can be reduced by incorporating the effects of conduction into the single-cell stimulation current.
We analyzed the usual monophasic and biphasic pulses and a new stimulation waveform that resembles the current associated with conduction in tissue (denoted as adjusted waveform). Simulations were run using the ten Tusscher-Panfilov model. Control conditions were simulated by stimulating fiber and single-cell models for 100 cycles at 1 Hz. Electrophysiological markers were computed from the last AP. Relative differences between each marker obtained in single-cell and in fiber simulations were computed. The individual contribution of ionic conductances to AP markers was also assessed.
Under control conditions, the adjusted waveform produced the smallest difference between cell and fiber for all markers except for APD25. In comparison with monophasic stimulation, the adjusted stimulation reduced cell-fiber differences by 99% for triangulation, 95% for Vmax, and by 76% for the maximum voltage slope. Regarding ionic contributors, notable differences were found for triangulation, where monophasic stimulation overestimated the contribution of INa and underestimated the contribution of ICaL. The adjusted waveform produced the best match between ionic contributors in cell and in fiber for most markers.
Our results indicate that the adjusted waveform decreases the mismatch between cell and fiber simulations, which could improve the trustworthiness of single-cell simulations without adding computational cost.