Medical decision-making diagnostic processes are based on collected data and knowledge, their analysis, and different ways of reasoning. In expert systems different methodologies are used to formalize the reasoning processes using gathered data and knowledge. Considering the support of diagnostic and therapeutic decision-making, attention is restricted to systems which would come into play when or after the patient data is collected. Electronic information sources such as hypertext systems and medical databases can be used to assist in decision-making, but they are not specifically designed to do so. It must be noted that technology itself cannot (yet) generate the knowledge base and identify what are the significant features required for the confirmation of a diagnostic situation. Specification of rules for identification can be achieved in other ways. Mostly clinical experts can express the factors and the interactions that they feel are clinically significant to identify a particular clinical condition. As clinical resources become more and more scarce, and the necessary competencies become even more in short supply, the expert system will act as a tool for practice guidance and also "insurance" when more junior staff, or those who do procedures less frequently, have to undertake diagnostic procedures that they are not fluent in. So for specific training, the reference and ongoing supervision expert systems, which reflect clinical "good practice" and leading edge-knowledge, will be increasingly useful. Resources, which need to be input into the current mechanisms for expert systems and artificial intelligence, will ultimately migrate from being research exercise to being cost-effective operational support both diagnostic support and clinical intervention activities.
Decision support systems in medicine should help physicians support their medical decision-making in an interactive way. They can be used in many fields, e.g. in diagnostic and prognostic decision-making, interpretation of laboratory tests, therapy planning, monitoring of treatment and diseases as well as in clinical management. However, if we are trying to design an "intelligent" system we are facing the problem how to model the intelligent behavior of human decision-making. The broad field that is now referred to as artificial intelligence deals with the problems that have until recently only been able to be tackled by humans because their formulation and solutions require some abilities that only exist in humans (such as the ability to memorize, think, observe, learn, see and similar senses). To these belong problems such as speech and pattern recognition, chess playing and diagnostic, therapeutic and prognostic medical decision-making. Turing  proposed an interesting test to find out if a computer exhibits intelligent behavior. His proposal was: "A computer could be considered to be thinking only when a human interviewer, conversing with both an unseen human being and unseen computer, could not determine which is which." This definition of artificial intelligence was focused on the comparison between the abilities of humans and the abilities of computers. Other definitions of artificial intelligence focus on decision-making and problem solving. In summary, the goal of artificial intelligence is to develop systems that behave intelligently. It aims to create computer hardware and software capable of emulating human reasoning.
The costs of hardware and software are declining whereas the capabilities of computer systems are continuously increasing. However, despite all the technological and methodological developments, many physicians or health managers are not using computers at all, or are using them primarily to support simple decisions. Decision support systems, especially knowledge-based systems and expert systems, are designed to change this situation. The classical definition of decision support system, provided by Keen and Scott-Morton, is: "A decision support system couples the intellectual resources of individuals with the capabilities of the computer to improve the quality of decisions. It is a computer-based support system for management decision makers who deal with semi-structured problems". As a special case of decision support systems we can consider knowledge-based or expert systems. We will work with the following definitions: "A knowledge-based system is a computer system that embodies knowledge, including inexact, heuristic, and subjective knowledge the results of knowledge engineering" . An expert system is defined as "a computer system that simulates a human expert in a given area of specialization" . Therefore an expert system can be seen as a knowledge-based system used for simulation of human expertise in a given area of specialization.
Medicine is the area where expert systems can rely on two basic types of medical knowledge : scientific knowledge (based on results of biomedical research) and empirical knowledge (based on experience gathered from diagnostic and treatment processes). Both types of knowledge are described in textbooks and other publications and especially scientific knowledge is taught at medical faculties. Scientific "know how" knowledge is of a cognitive type, i.e. it helps in recognizing the basis of biological processes, relationships among pathophysiological conditions and symptoms of diseases. Clinical experience is concentrated in medical documentation and it can be stored in medical databases. This empirical "know why" knowledge helps a physician to recognize a disease from observed features of a patient. In practice physicians consider both types of knowledge. Mostly physicians have sufficient scientific and empirical knowledge and no decision support systems are needed. However, there are situations when decision support systems are desirable. In expert systems different methodologies are used to formalize the reasoning processes using gathered data and knowledge.
Formalization and structuring of medical data and knowledge is not easy. Even in the case where we admit that all scientific and empirical knowledge is stored in computer we only can propose expert systems based on our currentmethodological achievements on how to make decision proposals. Till recently we have not known how the human brain process collects data and knowledge. In contrast to the human brain, decision-making using an expert system has been well described. Thus, the dream of the computer that performs at a high level of competence over a wide variety of tasks that people perform well seems to rest upon knowledge in the task areas.
In the paper by Feigenbaum  the knowledge principle was formulated as: "a system exhibits intelligent understanding and action at a high level of competence primarily because of the specific knowledge that it contains about its domain of endeavor". The knowledge principle simply says that reasoning processes of intelligent systems are generally weak and are not the source of power that leads to high levels of competence in behavior. Therefore one of the basic requirements to solve decision-making problems is to collect sufficient knowledge about it. The significance of research oriented to intelligent systems development for management and decision support in medicine and healthcare, including knowledge-based system, is increasing for the needs of the information society in healthcare.
Was this article helpful?