Brought to you by EarthWeb
IT Library Logo

Click Here!
Click Here!

Search the site:
 
EXPERT SEARCH -----
Programming Languages
Databases
Security
Web Services
Network Services
Middleware
Components
Operating Systems
User Interfaces
Groupware & Collaboration
Content Management
Productivity Applications
Hardware
Fun & Games

EarthWeb Direct EarthWeb Direct Fatbrain Auctions Support Source Answers

EarthWeb sites
Crossnodes
Datamation
Developer.com
DICE
EarthWeb.com
EarthWeb Direct
ERP Hub
Gamelan
GoCertify.com
HTMLGoodies
Intranet Journal
IT Knowledge
IT Library
JavaGoodies
JARS
JavaScripts.com
open source IT
RoadCoders
Y2K Info

Previous Table of Contents Next


5.5. ACCEPTABILITY ISSUES

Most systems have not been effectively utilized outside a research environment, even when their performance has been shown to be excellent. This suggests that it is an error to concentrate research primarily on methods for improving the system's decision-making performance when clinical impact depends on solving behavioral problems of acceptance as well.

There are some conditions which exist in clinical environment that are uniquely related to developing and using expert systems, which may result in "problems." In many instances, user expectations have been raised too far regarding what can be achieved from expert systems. Users are often expecting dramatic improvements in decision-making and innovative new ways of treating patients. One message is that users must have realistic expectations about what a system can do, how much it costs, and how long it will take to effectively use it. Furthermore, there are some who would say that the bias of medical doctors with respect to computers is so strong that systems will inevitably be rejected, regardless of performance (Startsman and Robinson, 1972). Training may be the best way to deal with these situations.

However, we are beginning to see examples of applications in which initial resistance to these systems has gradually been overcome through the incorporation of adequate system benefits. A heightened awareness of human engineering issues will also make systems more acceptable to physicians by making the system easier and more friendly to use. The issues range from the graphic user interfaces (GUI) capability of the system that tailors the style of the interaction to the needs and desires of individual physicians to the features of the system that make it appear as a helpful tool rather than a complicting burden. Introduction of the Intelligent Computer-Assisted Instruction (ICAI) feature to the medical expert system may be needed in designing the system because, according to the survey by Kraut and Mann in 1996, the most frequent users in the future are expected to be physicians with little computer knowledge.

6. FUTURE TRENDS

The applications of expert systems in medicine appear to be increasing at an almost exponential rate. However, among the expert systems that have been implemented, there are questions concerning the actual success of at least some of these implementations. Most of the systems published in papers have not been successful in practice, especially in the clinical environment. In some cases, failures have definitely occurred, and many of these failures have been due to an improper selection of domains or a neglect of the critical factor of expert system maintenance. In others, failure may be traced to the choice of the wrong knowledge acquisition and representation methods. However, several studies have shown that most problems encountered in the implementation of expert systems have not been a fault of the methodology. Rather, the faults most often can be ascribed to ethical and legal problems, behavioral problems, and certain errors on the part of those who have attempted to implement the methodology. Current research suggests that the following areas require attention in the future.

6.1. PROCESSING OF TEMPORAL KNOWLEDGE

Physicians faced with diagnostic and therapeutic decisions must reason about clinical features that change over time. In diagnostic, prognostic, and therapeutic problem-solving, the chronology of disease symptoms, cures, and failures gives the physician insight into what may lie ahead for their patient. Temporal reasoning involves using a theory of change to predict the future or an explanation of the pat or present. Most current expert systems lack the ability to represent or infer temporal events consistently. However, a system that provides advice regarding patients with chronic diseases must be able to handle temporal issues since they should be managed with repetitive trials of empirical therapy.

Despite the importance of temporal features in medical decision-making, representing and reasoning about temporal concepts have proven difficult for medical expert systems. Simply adding a time stamp to each patient observation is insufficient in many applications. The appropriate interpretation of medical data is complicated by the effect of the current and past clinical contexts on the usefulness and validity of a measurement.

Temporal reasoning has been applied to the following areas in recent years: management of cholesterol (Rucker et al., 1990), interpretation of ECG (Tong et al., 1993), and patient monitoring (Shahar et al., 1993). As the prevalence of chronic diseases increases in the future, there will be many studies on temperoal reasoning that perform well in working medical expert systems. According to the 1996 survey conducted by Kraut and Mann, the most important future interests of research in the field of medical expert systems are: processing of uncertain and temporal knowledge (54%), aspects of system integration (35%), and aspects of knowledge acquisition and maintenance (27%).


Previous Table of Contents Next

footer nav
Use of this site is subject certain Terms & Conditions.
Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details.