Computing cardiac recovery maps from electrograms and monophasic action potentials under heterogeneous and ischemic conditions

The currently available techniques to investigate the 3D sequence of activation and recovery in the cardiac atria and ventricles, with high spatial resolution, are based on extracellular electrical recordings. The goal of the present work is to provide an extensive quantitative analysis of the accuracy level of commonly used recovery time (RT) markers, under heterogeneous and pathological conditions of the myocardial tissue, such as myocardial ischemia. A widely used technique is based on unipolar electrograms (EGs); an alternative technique is based on hybrid monophasic action potentials (HMAPs), obtained as the potential difference between a permanently depolarized site and an exploring site. The RT markers derived from EGs and HMAPs are compared with two transmembrane action potential (TAP) markers considered here as gold standards for the fastest and final recovery phase, respectively. The analysis is based on 3D numerical simulations of the action potential propagation in anisotropic and insulated cardiac blocks, modeled by the Bidomain system coupled with the Luo\u2013Rudy I membrane model. These demanding simulations have been made possible by recent advances in computing power and multilevel Bidomain solvers. The results show that the extracellular RT markers considered are reliable estimates of the gold standard TAP markers, with low relative mean discrepancies and high correlation coefficients. We also investigate the capability of the markers to discriminate different transmural dispersions of recovery times and action potential durations. In some specific pathological cases when the EG markers fail, the HMAP markers may offer reliable alternatives