Deep learning based model for decision support with case based reasoning

Abstract

Cardiovascular diseases (CVD) the most common reason for deaths worldwide, are not easily diagnosed in initial stages. Early and accurate detection of CVD is highly required to prevent this leading cause of mortality. In the last few years, with the advancement of technology and increased potential of digital procedures, almost every business sector is adapting the automation and hence generating a large volume of data. Health sector is also affected by this outburst of technology and almost all the hospitals are generating a huge volume of data every day. The need of the hour is how to handle such a huge data and finding the hidden correlations among it so that it can be used by clinical experts in disease diagnosing and helps them in decision-making. This paper presents an intelligent decision support model for detection of coronary artery disease (CAD) with the integration of cuckoo algorithm for feature subset, analysis of various classification techniques to diagnose the disease more accurately and case base reasoning (CBR) for detecting the severity of the disease. The results seems promising and the integrated technique shows the accuracy of MLP is 85.48%.The model can be used as a promising decision making tool for medical experts for detecting cardio vascular diseases in their early stages.