Key result
Novel AdaBoost framework identifies CAD with ~94% accuracy, outperforming standard classifiers.
Why the study?
Accurate diagnostic classification and disease detection of coronary artery disease are crucial to aid clinical decision-making and reduce diagnostic errors in high-risk patients.
Does a genetic algorithm-based Improved AdaBoost Ensemble Classifier improve diagnostic classification accuracy of coronary artery disease compared to standard classifiers?
Cross-Sectional (n=1,779)
Yes
Does a genetic algorithm-based Improved AdaBoost Ensemble Classifier improve diagnostic classification accuracy of coronary artery disease compared to standard classifiers?
p-value: p=< 0.05
A novel framework integrating genetic algorithm-based feature selection with an Improved AdaBoost model demonstrated high accuracy for cross-sectional detection of coronary artery disease, outperforming standard machine learning classifiers.
Background Cardiovascular diseases (CVDs) remain the primary contributor to global morbidity and mortality, with coronary artery disease (CAD) representing the most prevalent and critical condition. Accurate diagnostic classification and disease detection are crucial for aiding clinical decision‐making and reducing diagnostic errors in high‐risk patients. Objective This study proposes a novel diagnostic classification framework that integrates genetic algorithm (GA)–based feature selection with an Improved AdaBoost Ensemble Classifier for cross‐sectional CAD detection. The primary goal is to develop a robust model where K‐fold cross‐validation and a voting mechanism are embedded directly within the boosting loop to generate diverse, weighted subclassifiers for enhanced generalization and classification accuracy. Method A public dataset (1190 records) and a local cohort collected from Noor Heart Center (589 records) were evaluated, with the positive clinical outcome strictly defined as confirmed CAD (presence of ≥50% vessel stenosis). GA was employed to identify the most informative clinical feature subsets by handling nonlinear interactions. Subsequently, the proposed Improved AdaBoost framework was applied for diagnostic classification, and its performance was benchmarked against various individual and ensemble classifiers. Results The GA successfully identified chest pain type (cp), ST depression (oldpeak), slope, and patient sex as the most predictive clinical features across both datasets. Using this optimal feature subset, the proposed framework achieved superior diagnostic performance. Specifically, the Improved AdaBoost model attained an accuracy of 93.25%, a precision of 95.31%, and an AUC of 95.52% on the public dataset, and an accuracy of 94.50%, a precision of 95.88%, and an AUC of 95.91% on the local dataset. This performance significantly outperformed standard AdaBoost and the top‐performing individual classifiers such as RPART ( p < 0.05). Conclusion The empirical findings confirm the potential of the proposed GA‐based Improved AdaBoost framework as a dependable, highly accurate decision‐support tool for cardiologists in the cross‐sectional diagnostic classification of CAD, facilitating timely patient management and clinical resource optimization.
No takes yet. Share an insight, caveat, or question.
Soleimandarabi et al. (2026) conducted a cross-sectional in Coronary artery disease (n=1,779). GA-based Improved AdaBoost Ensemble Classifier vs. Standard AdaBoost and individual classifiers (e.g., RPART) was evaluated on Diagnostic classification of confirmed CAD (p=< 0.05). A genetic algorithm-based Improved AdaBoost framework classified coronary artery disease with 93.25% to 94.50% accuracy and 95.52% to 95.91% AUC, outperforming standard classifiers (p<0.05).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: