Improving Diagnosis of Breast Cancer Disease Using Adaptive Neuro-Fuzzy Inference System

Cancer and particularly breast cancer is one of the most common diseases among women worldwide. Early detection of breast cancer is a major challenge for physicians and is key in successful treatment and patient survival. This study introduces some data mining methods for the prediction of breast cancer based on a dataset containing 683 independent records with 9 features from the UCI machine learning repository. The models were used to diagnose benign and malignant breast cancer. Results showed that the accuracy of Multi-Layer Perceptron Neural Network (MLP), Learning Vector Quantization (LVQ) Neural Network, Radial Basis Function (RBF), Fuzzy Clustering (KFC), Adaptive Neuro-Fuzzy Inference System Model (ANFIS) were 97.5%, 97.5%, 98.3%, 75% and 99.2%, respectively. Early diagnosis of breast cancer disease reduces the cost of treatment and increases the chance of successful treatment. This study demonstrated that neuro-fuzzy inference system performed better than other models for breast cancer diagnosis. In this study, while diagnosing breast cancer, it was illustrated that models based on fuzzy neural inference had a more acceptable performance than other methods in diagnosing breast cancer. The proposed model can assist the medical community, particularly mammography specialists.