Comparison of coronary artery disease guidelines with extracted knowledge from data mining

Area of Science:

• Cardiology
• Data Mining
• Health Informatics

Background:

• Coronary artery disease (CAD) is a leading global cause of mortality and disability.
• Up-to-date national guidelines are crucial for managing heart-related diseases effectively.
• Data mining offers a powerful approach to uncover insights from patient data.

Purpose of the Study:

• To identify significant risk factors and diagnostic patterns for Coronary Artery Disease (CAD) using data mining.
• To compare data mining-derived knowledge with established Canadian Cardiovascular Society (CCS) and US National Institute of Health (NIH) guidelines.
• To rank the importance of risk factors and diagnostic methods based on local patient data.

Main Methods:

• Analysis of 1993 completed patient records with CAD from 2009-2014.
• Utilized 25 variables, including CAD status and 24 predictor variables, for knowledge discovery.
• Compared extracted data mining rules with CCS and NIH guidelines, ranking findings by importance.

Main Results:

• Chest pain was identified as the most significant factor influencing CAD.
• Elderly males (age >54) showed a high probability of CAD diagnosis.
• Blood tests demonstrated greater diagnostic value compared to other recommended medical tests.

Conclusions:

• Data mining techniques confirm and enhance existing CAD guidelines by ranking risk factors.
• Extracted knowledge revealed novel patterns in CAD patient data.
• The study validates the utility of comparing data mining results with established clinical guidelines.