A rough set-based reasoner for medical diagnosis

Diagnosis of breast cancer analysis disease becomes one of an open discussion and a crucial need in Egypt. The analysis of these datasets for patients is important for the early detection and prediction of the disease. The usage of case-based reasoning (CBR) systems and the machine learning techniques provides us with several techniques to easily decide whether the patient is healthy or not. In this paper, we proposed a case- based reasoner architecture that aid physicians to early detect and predict breast cancer disease. As a retrieval technique Rough Sets Theory (RST) is applied followed by two different classifiers to improve the classification accuracy of the medical data. Results yield to 96% accuracy for 103 out of 108 instances and 82% classification accuracy after the usage of two different classifiers other than the RST (Neuro-Fuzzy and K-Nearest Neighbor classifiers).