By creating a monitoring system to detect potentially fatal physiologic trends, we can provide physicians with an automatically-generated differential diagnosis based upon a patient's recent history. In order to make this diagnosis clinically reliable, the system must be sensitive enough to discriminate between physiologic events that can share nearly identical trends. In an effort to refine our discriminatory techniques, we have created a software "monitor" that uses fuzzy logic to analyze physiologic signals to make a clear distinction between an arrhythmic cardiac arrest (ACA) and the onset of cardiopulmonary bypass (CPB). The differentiation capabilities of this monitor will form the foundation for a comprehensive automated diagnostic system.