Detection of Cardio Vascular Disease Using Fuzzy Logic

Abstract

The aim of this work is to detect whether an individual has a Cardio Vascular Disease (CVD) or not by using Mamdani and Sugeno methods of Fuzzy Inference System (FIS). The data set used for this work consists of 1000 records from the pathology reports of Thyrocare, Suburban Diagnostics, Medall, SRL Diagnostics and Metropolis. The parameters considered for predicting whether the individual has CVD or not are blood pressure, blood sugar, heart rate and oxygen level in the blood (SPO2). The FIS outputs indicate (1) whether the individual has a CVD or not, (2) the risk level of CVD and (3) some primary level precautions depending on the risk level. Mamdani and Sugeno methods were evaluated by comparing them with the results of the pathology reports. The results show that the Sugeno method gives 2% more accuracy in predicting a CVD as compared to the Mamdani method. Sugeno FIS gives more dynamical values as compared to Mamdani FIS for different values of input which leads to higher accuracy of Sugeno FIS.