Multimodal Comparison of Left Ventricular Function and Volume Assessment Methods in Patients With Coronary Artery Disease: Echocardiography, Nuclear Myocardial Perfusion Imaging, and Cardiac Magnetic Resonance

ABSTRACT Background To evaluate the correlations and agreement between echocardiography, nuclear myocardial perfusion imaging (MPI), and cardiac magnetic resonance (CMR) for assessing left ventricular (LV) function and volumes in patients with coronary artery disease (CAD). Methods Forty‐eight patients with suspected or confirmed CAD who underwent echocardiography, nuclear MPI, and CMR within 1 week were included. The LV ejection fractions (EF), end‐diastolic volume (EDV), and end‐systolic volume (ESV) obtained using each modality were compared using correlational and Bland–Altman analyses, with CMR serving as the reference standard. Additionally, 13 N‐NH 3 and 99m Tc‐sestamibi ( 99m Tc‐MIBI) MPI subgroup analyses were conducted using CMR as the reference. Results Nuclear MPI showed better correlation and agreement with CMR than echocardiography for EF values ( r = 0.93 vs. 0.88; bias: −2.8% vs. 4.4%). For LV volumes, nuclear MPI and CMR showed excellent correlations (EDV/ESV: r = 0.94) with better ESV agreement (bias: −10.1 mL) than EDV (−25.4 mL). In subgroup analyses, 13 N‐NH 3 PET MPI maintained excellent correlations with CMR (EF: r = 0.95; EDV/ESV: r = 0.97), with minimal EF bias (−0.7%). 99m Tc‐MIBI SPECT MPI showed a greater EF bias (−7.8%) despite a good correlation ( r = 0.95) and wider limits of agreement for LV volumes (EDV: −92.3 to 47.4 mL; ESV: −66.0 to 61.7 mL). Conclusions 13 N‐NH 3 PET MPI demonstrates excellent correlation and agreement with CMR, providing a reliable alternative for assessing LV function and volumes in patients with CAD. Contrarily, echocardiography systematically overestimates EFs. These findings provide practical insights into discrepancies among imaging modalities and may support clinical decision‐making in CAD management.