The major advances in nonlinear science that have occurred over the last two decades are beginning to find significant illustrations in the electrophysiology of the heart. In particular, some of the electric characteristics of cardiac arrhythmias seem to belong to one or other of several well-explored universality classes of nonlinear dynamical behavior. Within each such class, the gross features of the process show considerable independence of the details. This suggests that, even with incomplete information, the suppression or management of arrhythmias could be accomplished by measures known to suppress corresponding phenomena in nonlinear dynamics. In the case of arrhythmias, these measures require the development of sensors and microprocessors controlling appropriate pacing action. such an artificial pacemaker should have commercial value at least on a par with that of the most sophisticated existing devices. Phase l is to consolidate current understanding of arrhythmia into computer software (and possibly a 'hard-wired' device) that mimics the underlying electrophysiology as closely as possible. Schemes for appropriate pacing devices will then be tested on these simulations. This should set the stage for Phase II, in which actual devices will be constructed and tested.
