Expert System Example

by Julie Herzner and Miriam Kubiska, Dec. 1992

A specific example of a expert system is PXDES which is a pneumoconiosis, a lung disease, X-ray diagnosis. This expert system incorporates the inference engine to examine the shadows on the X-ray. The shadows are used to determine the type and the degree of pneumoconiosis. This system also includes three other modes: the knowledge base, the explanation interface, and the knowledge acquisition modes. The knowledge base mode contains the data of X-ray representations of various stages of the disease. These elements are in the form of fuzzy production rules discussed in the previous paragraphs. The explanation interface details the conclusions, and the knowledge acquisition mode allows medical experts to add or change information in the system. Another example of a diagnosis rule-based expert system is EMERGE designed to be used in an emergency room. This system uses a form of production rules which incorporates weighing factors which are determined by a neural network. The neural network is composed of input and output blocks with a hidden layer block in between which communicates input to the output. The neural network learns from examples then predicts an output based on this knowledge. This system also uses an IF-THEN-UNLESS statement instead of an IF-THEN statement. Because of this, the decision process may be more precise, the results may be more accurate and the explanations may be better understood. A specific example from EMERGE using the IF-THEN and the IF-THEN-UNLESS indicates that the latter case is more compact and less complicated.

       	IF-THEN                                IF-THEN-UNLESS
IF forced volume capacity is high (.625)      IF forced volume capacity is high
AND Bronchoscopy results are positive (.250)  AND Bronchoscopy results are positive
AND local symptoms are present (.125)         AND local symptoms are present
THEN surgery is probably necessary            THEN Surgery is probably appropriate
IF Metastasis is present (.500)               UNLESS Metastasis is present
OR Contraindications to surgery exist (.500)  OR Constraindictions to surgery exist
THEN surgery is probably not appropriate

The IF-THEN-UNLESS rule allows for less rules, therefore, permitting a more rapid search. The logical operators, AND and OR, used in both these methods give their corresponding rules different weighing factors and can be seen in the parenthesis next to each statement in the IF-THEN case. The threshold value for OR is lower than that for AND. Thus, for each additional AND you have a lower weighing factor but for each OR, the weighing factor is divided evenly. It can be seen from the example above that the weighing factors must add up to one.