Spiral defect chaos with intermittency increases mean termination time.

Cardiac models are examples of excitable systems and can support stable spiral waves. For certain parameter values, however, these spiral waves can become unstable, resulting in spiral defect chaos (SDC), characterized by the continuous creation and annihilation of spiral waves and thought to underlie atrial fibrillation. During SDC, the number of spiral waves fluctuates and drops to zero at termination. In this work, we demonstrate that varying a single parameter allows the system to transition from SDC to a single spiral wave, passing through an intermediate regime of intermittency. In such intermittent dynamics, intervals of SDC are sandwiched between non-SDC intervals during which the number of spiral waves remains small and constant. We quantify this intermittency and show that the mean termination time increases significantly as the control parameter approaches values for which a single spiral wave is stable. In addition, we observe that quasistable spiral waves may intermittently persist in part of the computational domain, while the rest of the domain exhibits SDC. Our results may have implications for clinical atrial fibrillation, which often shows intermittency, switching back-and-forth between fibrillation and normal sinus rhythm.