There are several problems in representing medical knowledge in clinical routine which should be notified as follow:
Complexity of Medical Knowledge: Medical knowledge is inherently uncertain and vague. In addition, there are often large numbers of medical items; e.g. an ordinary medical decision-support system contains more than 2000 symptoms and about 400 diagnoses. Also, there are often large amount of relationships between medical items; such that the above mentioned medical decision-support system normally defines about 50000 associations between symptoms and diagnoses, between symptoms and symptoms, as well as between diagnoses and diagnoses.
Medical Knowledge Acquisition: Often several medical experts are necessary to develop medical knowledge bases. Though, in practice, it is a problem to motivate medical experts to represent medical knowledge because of their lack of proficiency and also, interest. Moreover, medical experts are not frequently able to provide agreed-upon and precise definitions of many medical concepts and medical relationships. As a result, medical experts accept recommendations of a knowledge-based system on other experts’ knowledge only reluctantly.
Organizational and Technical Problems: Today’s data gathering methods and their capabilities are not so sufficient to build medical knowledge bases in an effective manner. For example, there are many disease profiles in a very narrow diagnostic area that they should be gathered in a medical decision-support system to be capable of assisting in diagnosis; so it will be a very time consuming task. To reduce the time, one promising approach is to provide a uniform and commonly accepted framework to represent medical knowledge. The problem is that such a framework is not available frequently. Also, it is often very difficult to convince different specialists to use one unique formal framework in their systems. This is mainly because of the reasons stated in the above said items and the lack of functionality, reliability, and correctness of available knowledge bases.
Fundamental Knowledge Representation Problems: The available medical knowledge is sometimes incomplete because of incomplete physiological models of the human body. Besides, since most kinds of medical knowledge are either established by definition, statistical utilization, or by a physiological model, some sort of inconsistency may result when several experts represent medical knowledge in the case of last item. Also, the representation and processing of temporal knowledge, especially in the field of uncertain and vague information, is inherently a difficult task.
In spite of the problems stated above, there are many motivations for representing knowledge in medicine including medical diagnosis assistance, education of students, training of physicians, research resources, and so on.