Magnetocardiography (MCG) has been proposed as a non-invasive and functional technique with high accuracy for diagnosis of myocardial ischemia.
This study sought to develop a novel scoring system of MCG for predicting the presence of significant obstructive coronary artery disease (CAD).
In a training set of 108 subjects, predictors of ≥70% stenosis in at least one major coronary vessel were prospectively identified from MCG variables. The final model was then retrospectively validated in a separate set of 45 subjects.
In the multivariable logistic regression, among those in the training set, elevated scores were predictive of ≥70% stenosis in all subjects (OR: 40.85; 95% CI: 6.28–265.90; p < 0.001). In the validation set, the score had an area under the receiver-operating characteristic curve of 0.91 (p < 0.001) for ≥70% stenosis. At an optimal cutoff, the score had 89% sensitivity, 77% specificity, 74% positive predictive value (PPV), 91% negative predictive value (NPV), and 82% accuracy for ≥70% stenosis. Partitioning the score into three levels of predicted risk, 91% of subjects could be identified or excluding CAD (81% PPV and 84% NPV).
We described an MCG score with high accuracy for predicting the presence of anatomically significant CAD.