Title: Design and Evaluation of a Fuzzy Expert System for Early Detection of Breast Cancer
Authors: Mazen S. Ihlayyel , Samy S. Abu-Naser
Volume: 9
Issue: 8
Pages: 116-120
Publication Date: 2025/08/28
Abstract:
This paper proposes a fuzzy expert system for early detection of breast cancer using linguistic variables derived from clinical and blood-based biomarkers. Two foundational studies were reviewed: one implementing a fuzzy inference system (FIS) based on patient-reported symptoms and expert rules, and the other utilizing biochemical parameters from the Coimbra dataset. The new hybrid model incorporates symptom-based and biochemical variables and uses Mamdani fuzzy inference to classify breast cancer as benign or malignant. The system was implemented in CLIPS and python () , and its accuracy was evaluated using a confusion matrix. Results showed high accuracy, sensitivity, and specificity, demonstrating the model's effectiveness in supporting breast cancer diagnosis. Final evaluation results showed an accuracy of 90.3%, sensitivity of 87.3%, and specificity of 95%.