Image Quality and Reliability of a Novel Dark-Blood Late Gadolinium Enhancement Sequence in Ischemic Cardiomyopathy

Purpose: The aim of this study was to assess the reliability of a 2D dark-blood phase-sensitive late gadolinium enhancement sequence (2D-DBPSLGE) compared with 2D phase-sensitive inversion recovery late gadolinium enhancement sequence (2D-BBPSLGE) in patients with ischemic cardiomyopathy (ICM). Materials and Methods: A total of 73 patients with a clinical history of ICM were prospectively enrolled. The following endpoints were evaluated: (a) comparison of image quality between 2D-BBPS(LGE)and 2D-DBPS(LGE)for differentiation between blood pool-late gadolinium enhancement (LGE), remote myocardium-LGE, and blood pool-remote myocardium; (b) diagnostic accuracy of 2D-DBPS(LGE)compared with gold standard 2D-BBPS(LGE)for the evaluation of infarcted segments; (c) diagnostic accuracy of 2D-DBPS(LGE)for the evaluation of microvascular obstruction (MVO); (d) comparison of transmurality index between 2D-BBPS(LGE)and 2D-DBPSLGE; (e) comparison of papillary muscle hyperenhancement between 2D-BBPS(LGE)and 2D-DBPSLGE; inter-reader agreement for depiction of hyperenhanced segments in both LGE sequences. Data were analyzed using pairedttest, Wilcoxon test, and McNemar test, and eta(2)coefficient and intercorrelation coefficient (ICC). Results: Image quality was superior for 2D-DBPS(LGE)for differentiation of blood pool-LGE (P<0.001). 2D-DBPSLGE, compared with 2D-BBPSLGE, showed a sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of 96.93%, 99.89%, 99.71%, 98.78, and 99.04%, respectively. Concerning MVO detection, 2D-DBPS(LGE)showed a sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of 66.67%, 100.00%, 100.00%, 80.95%, and 86.21%, respectively. 2D-DBPS(LGE)underestimated the transmurality (P=0.007) and identified papillary muscle hyperenhancement (P<0.001). Both LGE sequences showed comparable interobserver agreement for the evaluation of infarcted areas (2D-BBPSLGE: ICC 0.99;2D-DBPSLGE: ICC 0.99). Conclusions: Compared with 2D-BBPSLGE, 2D-DBP(SLGE)sequences provide better differentiation between LGE and blood-pool, while underestimating LGE trasmurality and the presence of MVO.