Information Technology Reference
In-Depth Information
medical systems are complex systems involving inexact, uncertain, imprecise and
ambiguous information [15].
Fuzzy Cognitive Maps have been successfully used to develop Medical Decision
Support System (MDSS). A specific type for Medical Diagnosis is the Competi-
tive Fuzzy Cognitive Map (CFCM) [14, 15] which consists of two main types of
concepts: diagnosis-concepts and factor-concepts. Fig. 27.2 illustrates an example
CFCM model that is used to perform medical diagnosis. Here the concepts of the
FCM and the causal relations among them that influence concepts and determine
the value of diagnosis concepts indicating final diagnosis are illustrated.
Fig. 27.2
A CFCM model for Medical Diagnosis
In the CFCM model each diagnosis concept represents a single diagnosis, which
means that these concepts must be mutually exclusive because the main intention is
to infer always only one diagnosis. This is the case of most medical applications,
where, according to symptoms, medical professionals conclude to only one diagno-
sis and then they decide accordingly the most appropriate the treatment. Actually,
this comes from the medical axiom: “every patient has only one disease” but may
represent many symptoms related to different diseases but all are results of the prim-
itive disease. The general diagnosis procedure is a complex process that has to take
under investigation a variety of interrelated factors, symptoms and functions. Usu-
ally, in any real world diagnosis and decision problem, many different factors have
to be taken into consideration to conclude the most appropriate diagnosis. In accom-
plishing any diagnosis process, some of these factors are complementary, others are
similar and even others are conflicting.
In the Competitive Fuzzy Cognitive Map model, the factor-concepts represent
inputs into the MDSS, their values correspond topatient data, observed symptoms,
