Diagnostic performance of cardiac magnetic resonance imaging in coronary artery disease

Background: Cardiac magnetic resonance imaging (CMR) is a promising method for detecting coronary artery disease (CAD). The first reports of new diagnostic techniques indicated generally unrealistic diagnostic performance relying on retrospectively observed cut‐off values of quantitative parameters. Although visual analysis of CMR is the most applicable method for clinical work, its diagnostic performance is not fully elucidated for study components such as wall motion, perfusion and late enhancement in patients with different severity of CAD. Methods: A total of 30 subjects including 20 patients with CAD and 10 healthy volunteers were selected for the study. Of the patients, ten had stable CAD, five confirmed myocardial infarction (MI) without Q‐waves in electrocardiogram (ECG) and five confirmed MI with Q‐waves in ECG. All patients underwent coronary angiography and CMR for evaluating resting wall motion, rest and stress perfusion and late enhancement. Results: Combining the data from the three CMR techniques, 12 out of 20 patients were correctly identified as having CAD, and all controls were found to be healthy. Sensitivity, specificity, accuracy, positive and negative predictive values were 60·0%, 100·0%, 73·0%, 100·0% and 55·6%, respectively. Of the CMR components, resting wall motion and late enhancement gave the most diagnostic yield. Conclusions: We conclude that evaluation of CAD is feasible in patients with different severity of CAD using visually analysed CMR, especially when available CMR methodologies are combined together.