Knowledge Acquisition for Diagnosis of Skin Diseases as an Initial Platform for an Expert System

摘要

Background: The diagnosis of skin diseases, especially in patients suffering from more than one disease or having similar symptoms, is very complex and access to the knowledge of skin diseases makes the design of an expert system easier. This research aimed to design a knowledge base used for diagnosis of complex skin diseases, selected by experts. Methods: This applied developmental research was conducted in 2015. The study population included 10 dermatologists of Razi Hospital, affiliated to Tehran University of Medical Sciences. Data collection was conducted through a questionnaire and a checklist. The questionnaire had face and content validity and was based on Likert scale according to the twelfth chapter of the International Classification of Diseases (Tenth revision). The questionnaires were administered to participants and collected after completion. A checklist of knowledge acquisition was designed for each disease based on the semiology book of skin diseases with “agree-disagree” options and completed by interviews. Signs and symptoms had an agreement with at least 70% of the experts, and symptoms that were added according to the experts’ proposal entered the checklist and was given to experts for consensus in future evaluations. The software used in this research was Clementine and its statistical method used was Stata. The data were analyzed using SPSS, 16. Results: The diseases including pemphigus vulgaris, lichen planus, basal cell carcinoma, melanoma, and scabies were selected to design the expert system. Confirmed signs and symptoms of the diseases selected by the experts included 106 causes. Conclusion: The choice of the selected diseases needed by specialists in the knowledge system is a very vital component needed in designing the expert knowledge base system to meet international standards based on international classification and according to the needs of specialists.