In the field of medicine and nursing, there are plenty of situations when doctors or nurses have to make decisions, prescribe medicines or define diagnosis when there is no sufficient information for selecting a certain variant, or where some factors are unknown. In such situations medial professionals rely on their experience and intuition. However, such decision-making process is quite risky, because human beings might make mistakes, forget something, may not operate large amounts of data quickly etc. Such human factors may influence health and life of the patients.
In order to help doctors, nurses and other medical professionals, computer based systems for making such decisions and analyzing information have been designed. The two main types of automated systems are decision making systems and expert systems. The aim of this essay is to discuss these two kinds of systems, analyze their applications in medicine or nursing, set examples of such systems and consider the dilemma of decision making in medical field.
An expert system is software that attempts to reproduce the performance of one or more human experts, most commonly in a specific problem domain, and is a traditional application and/or subfield of artificial intelligence (Chytil & Engelbrecht 1997). In traditional computer programs, data is analyzed and after computations result is show to the user. In expert systems, the process of decision-making is based on the data given to the system (which usually represents the predetermined base of human expert knowledge in the chosen area). Generally, the expert system consists of the knowledge base, also known as heuristics, and the inference engine. The inference engine makes associations between the facts using hierarchical relationships assigned to data elements. Expert systems might be used as separate decision-making systems, but more often they are aimed to assist the doctor in analysis of a problem where uncertain factors are included.
The most frequently mentioned expert system – Mycin – was developed in the early 70s at Stanford University. The system chooses the best way of treating the patient who probably has an infectious disease; the system identifies the virus and suggests treatment. In 1980s another expert system has been introduced – Internist 1. This system is designed to aid the physician to implement a differentiated diagnostics process (Chytil & Engelbrecht 1997).
Decision support systems
Decision support systems (DSS) are a specific class of computerized information systems that supports business and organizational decision-making activities. A properly-designed DSS is an interactive software-based system intended to help decision makers compile useful information from raw data, documents, personal knowledge, and/or business models to identify and solve problems and make decisions (Jones & Patronis 1996).The key to decision support systems is to collect data, analyze and shape the data that is collected and then try to make sound decisions or construct strategies from analysis.
Basically, a decision support system includes a dynamic knowledge base and inferencing mechanism; medial-logic modules are used for choosing the best solution. The DSS may be based on expert systems, on artificial neural networks, or include elements of both systems, which is called connectionist expert systems (Ifeachor 1998). Artificial neural networks have proven to be quite efficient for non-linear data modeling and decision-making. Neural networks imitate the actions of a human brain in the process of selecting a solution. The decision support systems represent wider class of applications because they may be used for making dynamic solutions and applying them in situations with changing environment (though proper training should be done to the system previously). Classical examples of decision support systems are clinical decision support systems such as AAPHelp, designed to support the diagnosis of acute abdominal pain and, based on analysis, the need for surgery; PIP (Present Illness Program) that gathered data and generated hypotheses about disease processes in patients with renal disease, ONCOCIN – a rule-based medical expert system for oncology protocol management designed to assist physicians with the treatment of cancer patients receiving chemotherapy (Armoni 2000). The latter system used a special language for modeling decisions and sequencing actions over time.
Decision making dilemma
At first glance, expert and decision making systems are ideal solution for medicine and should be used for diagnosing, prescribing treatment and solving problematic situations. However, there are several ethical and logical dilemmas concerning artificial intelligence in medical sphere.
First of all, there is a significant group of patients who would not trust computer systems and prefer to contact with a human being during the treatment. Secondly, the knowledge base and inferencing systems are programmed and taught by people, which means there might be ambiguous or mistaken patterns in the system. Thirdly, the changing environment and its estimation is in many cases still better done by people than by AI systems.
Taking into account all the above-mentioned facts, it is possible to conclude that the most efficient application of expert and decisions making systems is when close interaction between the system and the doctor is maintained. In case of collaboration experience and vision of the medical professional is supplemented by the ability of the AI system to analyze large amounts of data, speed and associative links.
Expert and decision support systems play an important role in the development of modern medicine; they help to eliminate human mistakes and significantly broaden the scope of analysis for doctors, nurses etc. However, human decisions cannot be totally replaced by automated systems and the best application of artificial intelligence in the medical sphere is when close collaboration between medical professional and expert or decision support system exists.
Armoni, Adi. (2000). Healthcare information systems: challenges of the new millennium. Idea Group Inc
Chytil, M. K. & Engelbrecht, R. (1997). Medical expert systems. Sigma Press
Ifeachor, Emmanuel. (1998). Neural Networks and Expert Systems in Medicine & Healthcare. World Scientific Publishing Company
Jones, Rebecca A. Patronis & Beck, Sharon E. (1996). Decision making in nursing. Cengage Learning