Alternative methods for the assessment of mechanical dyssynchrony using phase analysis of gated single photon emission computed tomography myocardial perfusion imaging

Abstract

Measurement of left ventricular (LV) mechanical dyssynchrony from single photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) allows optimization of cardiac resynchronization therapy in heart failure patients. We compared the discriminatory ability and reproducibility of a new software method, Corridor 4DM (4DM) to the established method, Emory Cardiac Toolbox (ECTb) in normals and heart failure patients. LV dyssynchrony was measured in 100 control (Group 1) and 100 patients with LVEF <35% (Group 2) using time to peak thickening with first harmonic, fourth harmonic, and volume curve methods with the 4DM software, and compared to ECTb. Of the 3 4DM methods, first harmonic had the best correlation with the ECTb (R = 0.88, slope = 1.00, P < 0.0001, bias = -0.18° [95% CI: -20°; 16°] for phase standard deviation; and similarly for histogram bandwidth, while volume curve analysis had the greatest variation. The intra and inter-observer reproducibility for 4DM time to peak thickening with first harmonic was very good (R = 0.99, P < 0.0001 and coefficient of variability 10% [95% CI 9.2-12%] for intra-observer, and R = 0.97, P < 0.0001, coefficient of variability 16% [15-17%] for inter-observer, respectively). Finally, in patients with LVEF <35%, the area under the curve on receiver operator characteristic analysis was 0.93 [95% CI: 0.89-0.97] to detect significant mechanical dyssynchrony (i.e. standard deviation ≥43°) using 4DM versus ECTb. The 4DM-software provides an accurate and reproducible alternative method of dyssynchrony analysis of SPECT MPI for evaluation and management of heart failure.