Simulating paced electrogram fractionation. Paced electrogram fractionation analysis (PEFA) may identify a re-entrant substrate in patients at risk of ventricular fibrillation (VF) by detecting prolonged, fractionated ventricular electrograms ("fractionation") in response to premature extrastimuli. Numerical simulations of action potential (AP) propagation through human myocardium following such premature stimulation were performed to study the relationship between electrogram fractionation, fibrosis, and changes in AP currents. Activation in a resistive monodomain 2 cm2 sheet of myocardium containing nonconducting fibrous tissue was modeled using standard numerical methods for solutions of partial differential equations using the Priebe-Beukelmann (PB) AP equations. Myocardial fibrosis significantly influenced electrogram morphology. High densities of closely spaced fibrous septa caused functional block and altered propagation paths at short coupling intervals, and produced large increases in electrogram duration similar to those associated with increased risk of VF in clinical studies. Prolongation of the cardiac AP using the heart failure variant of the PB model further increased the amount of fractionation and thereby replicated clinical recordings more closely than did fibrosis alone. Increasing AP dispersion by a variable reduction in the potassium current I(Kr) simulated results seen in patients with the long QT syndrome with an abrupt increase in electrogram duration, while a uniform reduction in I(Kr) alone did not result in fractionated electrograms. In contrast, increases in cytosolic Ca2+ and Ca2+ buffering by troponin to simulate HCM had little effect on fractionation. These results relate the effects of fibrosis, AP abnormalities, and dispersion of AP duration to the characteristic electrograms recorded in patients at risk of sudden death.