Quantification of late gadolinium enhancement cardiovascular MRI in patients with coronary artery chronic total occlusion. Issue 8 (August 2020)

• Record Type: Journal Article
• Authors: Zhang, L.J.; Dong, W.; Li, J.N.; Mi, H.Z.; Jiao, J.; Dou, R.Y.; An, J.; Liu, J.L.; He, Y.; Song, X.T.
• Is Part Of: Clinical radiology. Volume 75:Issue 8(2020)
• Page Start: 643.e19
• Page End: 643.e26
• Publication Date: 2020-08
• Subjects: Medical radiology -- Periodicals; Radiotherapy -- Periodicals; Radiotherapy -- Periodicals; Radiology -- Periodicals; Societies, Medical -- Periodicals; Medical radiology; Radiotherapy; Electronic journals; Periodicals; 616.0757
• DOI: 10.1016/j.crad.2020.03.032
• Languages: English
• ISSNs: 0009-9260

Abstract

Abstract : AIM: To determine the most accurate and reproducible semi-automated greyscale thresholding technique for quantifying late gadolinium enhancement (LGE) in cardiovascular magnetic resonance imaging (CMRI), by using positron-emission tomography (PET) as the reference standard in patients with coronary artery chronic total occlusion (CTO). MATERIALS AND METHODS: LGE in CMRI, single-photon-emission computed tomography (SPECT), and PET were performed within 1 week in each of 63 patients with known CTO. The presence and quantity of LGE were determined with greyscale thresholds of 2, 4, 5, 6, and 8 standard deviations (SDs) above the mean signal intensity for normal remote myocardium and full width at half maximum (FWHM). The infarcted myocardium was delineated by PET. RESULTS: Sixty-three patients and 1, 008 segments were analysed. Based on patient analysis, with PET as the reference standard, the 5 SD method yielded the strongest correlation ( r= 0.85, p< 0 . 0001) compared with the 2 SDs ( r= 0.42), 4 SDs ( r= 0.73), 6 SDs ( r= 0.81), 8 SDs ( r= 0.71), and FWHM ( r= 0.69; p< 0 . 001 for all comparisons). The 5 SDs threshold quantification showed high interobserver and intra-observer agreement (intraclass correlation coefficient [ICC]=0.90, p< 0 . 0001; ICC=0.93, p< 0 . 0001, respectively). CONCLUSIONS: Semi-automated LGE CMRI greyscale thresholding with 5 SDs above the mean signal intensity for normal remote myocardium yields the strongest correlation to the extent of LGE identified using PET and is highly reproducible in patients with CTO. Highlights: LGE CMR with 5 SDs yields the strongest correlation with PET in CTO. LGE CMR with 5 SDs is highly reproducible in patients with CTO. LGE CMR with 5 SDs should be considered for quantifying LGE in patients with CTO.