[Paper Review] Diagnosis of Coronary Artery Disease Using Artificial Intelligence Based Decision Support System

Noor Akhmad Setiawan, Venkatachalam, Paruvachi Ammasai+1 more|arXiv (Cornell University)|Jul 6, 2020

Rough Sets and Fuzzy Logic12 references51 citations

TL;DR

A fuzzy decision support system for diagnosing coronary artery disease is built using Rough Set Theory to extract rules, fuzzified via discretized attributes, and validated on UCI and hospital data, reportedly outperforming some clinical methods.

ABSTRACT

This research is about the development a fuzzy decision support system for the diagnosis of coronary artery disease based on evidence. The coronary artery disease data sets taken from University California Irvine (UCI) are used. The knowledge base of fuzzy decision support system is taken by using rules extraction method based on Rough Set Theory. The rules then are selected and fuzzified based on information from discretization of numerical attributes. Fuzzy rules weight is proposed using the information from support of extracted rules. UCI heart disease data sets collected from U.S., Switzerland and Hungary, data from Ipoh Specialist Hospital Malaysia are used to verify the proposed system. The results show that the system is able to give the percentage of coronary artery blocking better than cardiologists and angiography. The results of the proposed system were verified and validated by three expert cardiologists and are considered to be more efficient and useful.

Motivation & Objective

• Motivate the development of an AI-based decision support tool for coronary artery disease (CAD) diagnosis.
• Develop a fuzzy decision support system (FDSS) using rules extracted by Rough Set Theory.
• Discretize numerical attributes and fuzzify the extracted rules to enable probabilistic reasoning.
• Assign weights to fuzzy rules based on their support to reflect their significance.
• Validate the system on multiple CAD datasets including UCI data and hospital data, with expert cardiologist review.

Proposed method

• Construct a fuzzy decision support system for CAD diagnosis.
• Extract rules using Rough Set Theory from CAD datasets.
• Discretize numerical attributes and fuzzify the rules.
• Weight fuzzy rules according to their extraction support.
• Validate the system against datasets from UCI and Ipoh Specialist Hospital with expert input.

Experimental results

Research questions

• RQ1 Can a fuzzy decision support system using Rough Set-derived rules accurately diagnose coronary artery disease?
• RQ2 Do fuzzified rules with weights improve diagnostic decision quality compared with traditional methods?
• RQ3 How does the proposed system perform on UCI CAD datasets and real-world hospital data?
• RQ4 Is expert cardiologist validation consistent with the system’s outputs?

Key findings

• The system provides a percentage assessment of coronary artery blocking.
• Validation by three expert cardiologists supported the approach.
• The proposed system is described as more efficient and useful for diagnosis.
• Datasets include UCI data from the U.S., Switzerland, and Hungary, plus Ipoh Hospital data.
• The results are claimed to outperform traditional methods such as cardiologists and angiography in certain aspects.