Diagnosis of hypothyroidism using a fuzzy rule-based expert system

Background: Early and accurate diagnosis of many diseases is critical to their treatment. Today, the classification models based on fuzzy intelligent systems help the uncertainty conditions in medicine, as well as the classification of diseases. The main goal of this study is to diagnose the most common thyroid disorders, hypothyroidism, using a fuzzy rule-based expert system. Methods: In this study, the data from patients who referred to Imam Khomeini Clinic and Shahid Beheshti Hospital in Hamadan west of Iran were collected. The data contain 305 subjects in three classes which are normal, subclinical hypothyroidism and hypothyroidism. Collected variables include demographic, symptoms as well as laboratory tests. In order to the diagnosis of Thyroid disorders, a fuzzy rule-based classifier was designed. Predictive performance of this model was compared with a multinomial logistic regression model in terms of the accuracy and the area under a ROC curve (AUC) as well as sensitivity and specificity. Results: The results showed that designed fuzzy rule-based system works well in thyroid disorder prediction with about 97% accuracy. Also, the fuzzy classifier has a better performance than the logistic regression model, especially for the subclinical hypothyroidism class. Conclusions: Fuzzy rule-based classifier by utilizing overlapping sets has improved the efficiency of classification and decision-making systems. Also, by providing the possibility of using linguistic variables in the decision-making process, easier interpretation, can be used by doctors who are not familiar with modeling concepts.