COMP3330: History Assignment

Artificial Intelligence Applications In Medicine

Other Approaches

The following approaches are discussed below:

Case-based reasoning

Fuzzy logic

Natural language processing

Intelligent agents

Case-Based Reasoning

Case-based reasoning has not been used extensively in medical AI. However, one example comes from the Department of Systems and Computer Engineering, Carleton University, Ottawa.

A case-based reasoner tool has been developed, allowing users to compare the ten-closest matching cases to the newest patient admission, from a database of intensive care medical records. This is used in conjunction with a back-propagation, feed-forward artificial neural network that has been trained and tested to estimate patient outcomes (duration of artificial ventilation, length of stay, and mortality).

Fuzzy Logic

Fuzzy logic has been used successfully in medical applications, usually in conjunction with other AI techniques such as artificial neural networks. Examples include:

Ambulatory blood-pressure monitors:
These are produced by many firms, such as DynaPulse by Pulse Metric, and Omron by Omron Healthcare Inc, and use pattern recognition and fuzzy logic algorithms to increase measurement accuracy.
 

Automated BIS Controller:
Aspect Medical Systems has developed monitors to assess the depth of the anaesthesia state based on the statistically derived Bispectral Index (BIS) reflecting the level of sedation. The Automated BIS Controller, in development at the University of Pittsburgh Medical Center, has been successfully employed throughout community hospitals in Indianapolis.
 
The system uses a fuzzy logic control program to control the rate of anaesthetic drug infusion, improving the administration of anaesthesia during surgery and contributing to improved patient care at reduced cost.

Natural Language Processing

In the medical industry, natural language processing is being utilised to help handicapped people to perform common tasks, such as write an e-mail, open doors on command, etc.

As this technology develops, it has an even greater ramification in all medical AI products. For example, doctors will be able to communicate with the computer AI system without having to use their hands. AI technology will be able to assist the doctor with critical, condensed information as he/she performs operations and other important tasks.

Hundreds of natural language processing projects are currently underway throughout the world; however, none have been developed specifically for medical applications, and hence there is no discussion of such applications presented.

Intelligent Agents

Intelligent agent systems are one of the newest research areas in artificial intelligence. Over the past five years there has been a growing interest in the application of agent-based systems in health care.

Current topics of research include:

Communication and co-operation between intelligent agents:
To improve patient management (e.g. distributed patient scheduling)
 

Agents that provide remote or elderly care delivery:
To improve access to health care services
 

Multi-agent systems:
To assist doctors in the tasks of patient monitoring and diagnosis, and improve medical training and education (e.g. tutoring systems)
 

Medical agent-based decision support systems:
To assist in decision making
 

Information agents:
To gather, compile and organise medical knowledge available on the Internet

A lot of methodological and technical problems associated with the real deployment of health-care agent-based systems are beginning to be discovered by researchers in the field. These include:

regulations concerning privacy of medical data, and security of the transfer of patient information between agents

social acceptance of agent-based systems

lack of universally accepted standard medical ontology

lack of centralised control and communication standards

difficulty of integration with other types of software

legal and ethical issues related to the use of agents in health care

No specific examples of intelligent agents for medical applications could be located for inclusion in this report, and hence there is no discussion of such applications presented.