Cardiac Magnetic Resonance versus Single-Photon Emission Computed Tomography for Detecting Coronary Artery Disease and Myocardial Ischemia: Comparison with Coronary Angiography

Figure 1. Representative images from a 72 year-old male patient with chest pain. (A) Stress-rest cardiac magnetic resonance (CMR) images show stress-induced perfusion defects (ischemia) in the lateral-anterolateral and inferoseptal left ventricular walls of the basal, mid and apical slices (open arrows) and subendocardial late gandolinium enhancement (LGE) (non-transmural infarction) in the anterolateral wall at the basal and mid-ventricular slices (black arrows). (B) single-photon emission computed tomography (SPECT) shows a partially reversible defect (white arrows) at the lateral and anterolateral left ventricular walls, but fails to demonstrate the inferoseptal ischemia. (C) SCA demonstrates a significant (>70%) luminal stenosis (left panel, black arrow) at the obtuse marginal branch and occlusion of the right coronary artery (right panel, open arrow) confirming the CMR findings.

Figure 2. Receiver-operating characteristic (ROC) curves of summed stress score (SSS) for stress CMR and SPECT myocardial perfusion imaging (MPI) in detecting significant angiographic CAD (≥70% stenosis). The area under curve (AUC) was 0.78 and 0.59 for CMR and SPECT, respectively. Only CMR detected CAD had the ability to predict positive findings on coronary angiography (p < 0.01).

Figure 3. ROC curves of SDS (i.e., ability to detect stress-induced myocardial ischemia) for MPI by stress CMR and SPECT, as determined by coronary artery angiography (≥70% stenosis). The AUC was 0.82 for CMR and 0.67 for SPECT. Only CMR-determined ischemia had the ability to predict positive findings on coronary angiography (p < 0.01).