City Research Online

Abstract

This thesis illustrates the technology required to provide a new generation of clinical instrumentation systems for critical care medicine. This advance in measurement science is gained from the use of a knowledge-based component able to process information as well as data. To implement a clinical information system using knowledge-based technology requires prior knowledge of human and computer-based activity within the critical domain. A historical perspective is given to both of these topics which reflects the genesis of current practice. The application area is introduced by investigating a control system approach to managing patients who require ventilatory therapy.

It was found that no current methodology is wholly appropriate when a knowledge-based component is included in the technological paradigm. Therefore, a novel methodology for system design, implementation and evaluation is proposed, and its utility tested in the aforementioned application domain. The detailed processes involved in the evolution of a prototype system which aids the clinical user in the art of ventilatory therapy are shown. Three levels of machine intelligence are shown to be required, based on: context-sensitive deterministic mechanisms; pattern cognition; and decision support elements. A wider discussion of the important points raised in the practical use of the methodology focuses upon the philosophical basis of clinical information systems and the processes of knowledge elicitation, knowledge representation and intelligent system evaluation.

Publication Type:	Thesis (Doctoral)
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science R Medicine > RZ Other systems of medicine
Departments:	School of Science & Technology > Computer Science School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads