The tasks of monitoring and diagnosing intensive-care patients take knowledge and skill and demand correct action in complex, unexpected, and time-critical situations. Guardian, developed at Stanford University's Knowledge Systems Laboratory, is a knowledge-based system designed to perform these tasks for post-cardiac surgery patients.1 The system is an autonomous agent with a flexible architecture, in which several algorithms cooperate to produce diagnoses and treatment plans under hard real-time conditions. Guardian has undergone several tests, and, with the help of a patient-simulator system, we have com-pared its performance to that of human physicians. The results show that a system such as Guardian is clearly valuable as a decision-support tool in an intensive-care unit.

Barbara Hayes-Roth, Jan Eric Larsson, "Guardian: An Intelligent Autonomous Agent for Medical Monitoring and Diagnosis", IEEE Intelligent Systems, vol. 13, no. , pp. 58-64, January/February 1998, doi:10.1109/5254.653225