A speedy cardiovascular diseases classifier using multiple criteria decision analysis

Area of Science:

• Cardiology
• Medical Informatics
• Artificial Intelligence in Medicine

Background:

• Cardiovascular diseases (CVDs) cause 30% of global deaths annually.
• Aging populations and medical staff shortages exacerbate CVDs.
• Existing cardiovascular disease classifiers (CDCs) lack rapid evaluation capabilities.

Purpose of the Study:

• To develop a novel cardiovascular disease classifier (CDC) for rapid auto-diagnosis.
• To address the limitations of previous CDCs in terms of speed.
• To improve the efficiency of cardiovascular disease detection.

Main Methods:

• Incorporated analytic hierarchy process (AHP)-based multiple criteria decision analysis (MCDA).
• Developed feature vectors using a Support Vector Machine.
• MCDA was used for efficient patient weighting and feature selection.

Main Results:

• Successfully implemented a speedy detection of cardiovascular diseases.
• The new CDC utilizes the most meaningful features for accurate discrimination.
• Reduced the number of features required for classification through MCDA.

Conclusions:

• The developed CDC enables rapid and accurate cardiovascular disease diagnosis.
• This automated approach can help mitigate the impact of medical personnel shortages.
• The integration of AHP-MCDA offers a promising strategy for improving diagnostic efficiency.