Diagnosis of coronary artery disease
Pro Research Analysisby 
Searched over 200M research papers
Clinical History and Symptom Assessment in Coronary Artery Disease Diagnosis
The initial and most crucial step in diagnosing coronary artery disease (CAD) is a thorough clinical history and symptom assessment. Understanding the type, duration, and triggers of chest pain or angina is often more informative than physical examination or even electrocardiography, as up to 50% of patients may have a normal resting ECG despite having CADCandell‐Riera1994King2020. Evaluating risk factors, age, gender, and symptom patterns helps determine the likelihood of CAD and guides further diagnostic testing.
Noninvasive and Invasive Diagnostic Techniques for CAD
Traditional and Advanced Imaging Modalities
Noninvasive imaging techniques play a central role in diagnosing CAD. These include stress electrocardiography, stress echocardiography, single-photon emission computed tomography (SPECT) myocardial perfusion imaging, positron emission tomography (PET), coronary computed tomography angiography (CTA), and magnetic resonance imaging (MRI)Mangla2017Hong2004. These tests provide both anatomical and functional information about coronary arteries, helping to identify blockages and assess their impact on heart functionMangla2017Hong2004.
64-slice CT coronary angiography, in particular, has shown high sensitivity (≥90%) and specificity (88–90%) for detecting significant coronary stenosis, with negative predictive values up to 100%. However, positive findings often require confirmation with invasive coronary angiography, and patient selection is important due to radiation exposure risksStein2008Mangla2017Hong2004.
Invasive Coronary Angiography
Invasive coronary angiography remains the gold standard for confirming CAD, especially when noninvasive tests are inconclusive or when intervention is plannedMangla2017Hong2004. However, its use is increasingly reserved for cases where noninvasive imaging has already established a high likelihood of diseaseMangla2017Hong2004.
Machine Learning and Artificial Intelligence in CAD Diagnosis
Recent advances in machine learning (ML) and artificial intelligence (AI) have led to the development of automated, noninvasive diagnostic tools for CAD. ML algorithms such as support vector machines (SVM), random forests (RF), and deep learning models have demonstrated high accuracy in detecting CAD using clinical data and biomedical signals like ECG and heart rate variabilityGhiasi2020Garavand2022Dolatabadi2016+2 MORE.
For example, SVM and RF models have achieved area under the curve (AUC) values of 0.88 and 0.87, respectively, indicating strong diagnostic performance. Deep learning approaches, such as capsule networks applied to ECG signals, have reported diagnostic accuracies exceeding 99%. Decision tree-based models, like the classification and regression tree (CART), have also shown robust and fast predictions, outperforming several other ML techniques in classifying patients as normal or having CADGhiasi2020Alizadehsani2019.
Biomarkers and Emerging Diagnostic Approaches
In addition to imaging and ML-based methods, established and novel biomarkers are being explored for the diagnosis of chronic CAD. These biomarkers, combined with clinical assessment and imaging, may further improve diagnostic accuracy and patient stratification.
Conclusion
The diagnosis of coronary artery disease relies on a combination of clinical history, noninvasive and invasive imaging, and, increasingly, machine learning-based tools. Clinical assessment remains foundational, while advanced imaging and AI-driven models offer high accuracy and efficiency, supporting clinicians in making timely and cost-effective decisions. As technology evolves, integrating these approaches will continue to enhance the early detection and management of CADGhiasi2020Candell‐Riera1994Garavand2022+7 MORE.
Sources and full results
Most relevant research papers on this topic
Decision tree-based diagnosis of coronary artery disease: CART model
The CART model provides a simple and reliable diagnosis of coronary artery disease, with maximum accuracy, sensitivity, and specificity, outperforming other methods in identifying patients as normal or CAD.
Efficient Model for Coronary Artery Disease Diagnosis: A Comparative Study of Several Machine Learning Algorithms
SVM and RF algorithms are the most effective machine learning algorithms for diagnosing coronary artery disease based on patient examination data.
Automated diagnosis of coronary artery disease (CAD) patients using optimized SVM
The proposed algorithm accurately diagnoses coronary artery disease (CAD) from normal conditions using heart rate variability (HRV) signals from electrocardiograms, with a sensitivity of 98.43% and specificity of 100%.
Machine learning-based coronary artery disease diagnosis: A comprehensive review
Machine learning-based coronary artery disease diagnosis shows promise, but results vary significantly due to dataset characteristics and applied ML techniques, making it difficult to generalize results.
1D-CADCapsNet: One dimensional deep capsule networks for coronary artery disease detection using ECG signals.
The 1D-CADCapsNet model accurately detects coronary artery disease from ECG signals, providing a fast and accurate diagnostic tool for cardiologists.
DOI