City Research Online

Abstract

PONI is an interrelated system of computer hardware and software, which monitors patients in the operating rooms and performs related decision making functions. PONI gathers, analyses and collates physiological data about these patients. Its physiological monitoring function is used to chart the patient's state of ventilation, circulation, and all variables which may alter in life-threatening situations. PONI analyses such parameters as heart rate, blood pressure and respiratory gas content, and provides speedy notice, in the form of cautions and alarms when corrective action is needed. In current practice, the large number of false positive alarms signalling no harm to the patient, is a considerable problem. PONI addresses the problem by comparing and validating instrument readings in this situation involving a knowledge base whose core is a set of 46 rules. This information has been analysed statistically and has suggested the knowledge rules for monitoring of the patient population. The rules were applied to 25,507 warnings (20,471 cautions and 5,036 alarms) from 157 Day Surgery Unit patients undergoing 201 hours of surgery. Most of the cautions were validated by the system analysis, but 2,757 of the 5,036 alarms were invalidated, 1,082 were validated, and 1,197 left indeterminate. From a total of 25,507 alarms, the system validated 19,354, invalidated 3,284 and left in an indeterminate state 2,869. Thus, PONI points up a potential reduction from one alarm every 4 minutes to one every 16 minutes. It is planned to expand the knowledge rules of the prototype expert system seeking to maximise the efficacy of alarm information regarding patient management with the immediate aim of minimising the number of false positive alarms.

Publication Type:	Thesis (Doctoral)
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Departments:	School of Science & Technology > Department of Computer Science School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads