Novel CAD Diagnosis Method Based on Search, PCA, and AdaBoostM1 Techniques

Area of Science:

• Cardiology
• Medical Informatics
• Machine Learning

Background:

• Cardiovascular diseases (CVDs) are a leading global cause of mortality.
• Coronary artery disease (CAD) is a major type of CVD, necessitating early diagnosis for effective treatment.
• Existing CAD diagnostic methods often use numerous features, potentially hindering early detection.

Purpose of the Study:

• To develop a novel, highly accurate method for early CAD diagnosis.
• To reduce the number of features required for CAD diagnosis without compromising accuracy.
• To improve upon existing CAD diagnostic techniques through a combination of feature selection and machine learning.

Main Methods:

• A new CAD diagnostic method was developed, integrating eight search techniques, principal component analysis (PCA), and the AdaBoostM1 algorithm.
• The method utilizes only five specific features: age, hypertension, typical chest pain, T-wave inversion, and regional wall motion abnormality.
• The approach was tested on the benchmark Z-Alizadeh Sani dataset.

Main Results:

• The proposed method achieved a classification accuracy of 91.8% on the Z-Alizadeh Sani dataset.
• This accuracy represents the best performance reported to date on this dataset using a minimal feature set.
• The method demonstrated superior efficiency and classification performance compared to basic machine learning techniques and existing studies.

Conclusions:

• The developed method enables early and accurate diagnosis of CAD.
• Medical practitioners can adopt this approach to improve patient outcomes through timely CAD detection.
• The study highlights the effectiveness of combining advanced algorithms with a reduced feature set for medical diagnosis.