City Research Online

Abstract

This project is concerned with the development of intelligent agents for use in the Intensive Care Unit (ICU), based on the Multigraph Architecture (MA). The objectives of the work are twofold:

i. ) To consider the design issues of intelligent systems, using a model-based approach of adding intelligence to existing instrumentation and information systems, irrespective of their implementation.
ii. ) More specifically, to investigate the role of MA in the implementation of such systems.

The first objective will be achieved by writing a specification for an Integrated Model-based Development Environment (IMDE), identifying its functionality, user requirements, etc. The second objective is to implement a prototype of one such development environment. This will borrow concepts from a framework for intelligent process control for chemical plants, previously implemented in the MA.

Finally an ICU application will be developed, using this IMDE, to monitor the respiratory system. The benefits associated with using such a model-based framework include modular design, reusability of software components, automatic synthesis of run-time system from the models and simpler consistency and validation procedures. Another benefit that is of particular interest in patient monitoring systems is the integration of different software components. Because the MA is generic, it can be extended to support any modelling paradigm. This means that the development framework can handle all aspects in design of the system, e.g. instrument interface, user interface, signal processing and knowledge-based signal interpretation. To demonstrate this, the IMDE has been extended to include the Causal Probabilistic Network (CPN) modelling paradigm as an integral component. This facilitates the modelling of uncertaiilty.

Using the techniques described in summary above, the benefits gained from applying the multigraph architecture to intelligent patient monitoring can be ascertained. This forms the bulk of the work to be described in this thesis.

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