A Fuzzy Decision Support System for Diagnosis of Some Liver Diseases in Educational Medical Institutions

Abstract

Decision support systems improve medical diagnosis and minimize diagnostic errors. Existing diagnostic systems are often complex and exhibit limited performance on liver diseases, particularly the liver cancer. This paper presents a fuzzy decision support system for helping students diagnose some human liver diseases in educational medical institutions. The proposed system aims to improve real medical diagnosis processes. The approach has three basic steps: 1) symptoms-based diagnosis, 2) liver function-based diagnosis, and 3) image processing- based diagnosis. The proposed system employs two artificial intelligence techniques: fuzzy logic and image processing. The first is used for diagnosing liver diseases based on the liver function tests, while the second is used for diagnosing liver diseases such as the liver cancer, hepatitis, liver cirrhosis, liver fibrosis, and fatty liver. The proposed system combines two methods: the Mamdani inference and simulation method used in the MATLAB17 fuzzy logic toolbox, and the gray level co-occurrence matrix, for extracting the features of the second- order statistical texture of images acquired using computed tomography, magnetic resonance imaging, or ultrasound, for various liver diseases. Our results reveal a very good agreement between expert-made and system-made diagnoses, suggesting high accuracy.