Abstract
Over the past decade a number of attempts have been made at developing a computer-based system for the detection of arrhythmias, primarily PVC's (Premature Ventricular Contractions). The authors have attempted to overcome a number of the shortcomings by designing and implementing a system that relies on considerable human interaction in order to accomplish a quantification of ventricular arrhythmias. The interaction between personnel and the computer occurs on two levels: (1) Adjustments in the program parameters for an individual patient can be made; for example signal gain, slope aperture, large slope value for QRS detection, PVC width, etc. (2) After computer detection of significant events has occurred, they are digitally stored and can be displayed back to personnel for confirmation and information. In initial evaluation, 50 patients were studied. The computer correctly identified 99. 5% of all QRS complexes, and 94. 5% of all PVC's with only 4 false-positive PVC's out of almost 60,000 QRS complexes.