Application of overlapping areas of fuzzy triangles in a fuzzy MCDM model for lung disease classification

Abstract

This research proposes an innovative fuzzy Multi Criteria Decision Making (MCDM) approach for classifying lung diseases based on the overlapping areas of fuzzy triangular numbers. Seven clinically relevant symptoms—cough, productive cough, shortness of breath, fever, chest pain, weight loss, and night sweats—are modeled as fuzzy triangular membership functions and weighted according to expert derived diagnostic significance. By calculating the area of overlap between a patient’s fuzzy symptom vector and disease prototypes, the method generates a crisp similarity score through weighted aggregation, thereby handling the uncertainty inherent in medical data. The model was evaluated on a dataset of 15 common lung diseases, achieving an overall classification accuracy of approximately 87 % and correctly identifying pneumonia and tuberculosis as the most probable diagnoses with the highest crisp scores. The results demonstrate that incorporating fuzzy triangle overlap areas into an MCDM framework enhances both the objectivity and accuracy of lung disease classification, offering a transparent and interpretable tool for clinical decision support systems that can improve diagnostic performance in complex respiratory conditions.