Abstract: A non-invasive system and method of diagnosing and predicting cardiac disease in a patient's heart is disclosed that comprises a microprocessor which contains a signal processor and a pattern recognition processor; detect the electrophysiological signals of the heart whereby the signals are processed to create a pattern that represents the patient's heart. The pattern may be further processed by repeatedly comparing it to patterns stored within the pattern recognition processor whereby certain coronary diseases such as myocardial ischemia in the patient's heart may be diagnosed. During each heartbeat, at least a million different electrical signals are collected and the results of test are displayed on a screen. The results may include the diagnosis, computer generated image of the patient's heart identifying areas of any cardiac disease that has been detected and/or a two dimensional non-linear waveform representing the electrophysiological signals of the patient's heart.