Histopathological Validation of Dark‐Blood Late Gadolinium Enhancement MRI Without Additional Magnetization Preparation

• Published in: Journal of magnetic resonance imaging 2022-01, Vol.55 (1), p.190-197
• Main Authors: Holtackers, Robert J., Gommers, Suzanne, Heckman, Luuk I.B., Van De Heyning, Caroline M., Chiribiri, Amedeo, Prinzen, Frits W.
• Format: Article
• Language: English
• Subjects: Animal models; Animals; Bias; Blood; cardiac magnetic resonance; Contrast Media; Coronary artery; Correlation coefficient; Correlation coefficients; dark‐blood LGE; Female; Field strength; Gadolinium; Heart; histology; Histopathology; Image analysis; Image contrast; Image processing; In vivo methods and tests; Ischemia; late gadolinium enhancement; Magnetic Resonance Angiography; Magnetic Resonance Imaging; Magnetization; Myocardial infarction; Myocardium; Noise standards; Occlusion; Prospective Studies; Radio frequency; Statistical analysis; Statistical methods; Statistical tests; Swine; Ventricle
• ISSN: 1053-1807; 1522-2586
• DOI: 10.1002/jmri.27805

Background Conventional bright‐blood late gadolinium enhancement (LGE) cardiac magnetic resonance imaging (MRI) often suffers from poor scar‐to‐blood contrast due to the bright blood pool adjacent to the enhanced scar tissue. Recently, a dark‐blood LGE method was developed which increases scar‐to‐blood contrast without using additional magnetization preparation. Purpose We aim to histopathologically validate this dark‐blood LGE method in a porcine animal model with induced myocardial infarction (MI). Study Type Prospective. Animal Model Thirteen female Yorkshire pigs. Field Strength/Sequence 1.5 T, two‐dimensional phase‐sensitive inversion‐recovery radiofrequency‐spoiled turbo field‐echo. Assessment MI was experimentally induced by transient coronary artery occlusion. At 1‐week and 7‐week post‐infarction, in‐vivo cardiac MRI was performed including conventional bright‐blood and novel dark‐blood LGE. Following the second MRI examination, the animals were sacrificed, and histopathology was obtained. Matching LGE slices and histopathology samples were selected based on anatomical landmarks. Independent observers, while blinded to other data, manually delineated the endocardial, epicardial, and infarct borders on either LGE images or histopathology samples. The percentage of infarcted left‐ventricular myocardium was calculated for both LGE methods on a per‐slice basis, and compared with histopathology as reference standard. Contrast‐to‐noise ratios were calculated for both LGE methods at 1‐week and 7‐week post‐infarction. Statistical Tests Pearson's correlation coefficient and paired‐sample t‐tests were used. Significance was set at P