You are viewing the site in preview mode
Skip to main content
Disease and cohort name
Analysis method
N
LGE detection rate, n (%) (P value)
Scar amount (%), mean ± SD
Scar amount (%), median (IQR) (P value)
B-A plot mean ± LoA
CCC (P value)
Non-ischemic training cohort
ECV-guided LGE analysis versus Manual analysis
80
61 (76.3) versus 80 (100) (p < 0.01)
1.8 ± 2.9 versus 5.1 ± 3.0
0.92 (0.1–2.1) versus 4.5 (3.2–6.4) (P < 0.01)
− 3.2 ± 4.0
0.48 (P < 0.01)
Non-ischemic validation cohort
ECV-guided LGE analysis versus Manual analysis
20
10 (50.0) versus 18 (90.0) (p < 0.01)
1.1 ± 1.7 versus 2.9 ± 2.4
0.2 (0–1.6) versus 2.5 (1.2–3.7) (P < 0.01)
− 1.8 ± 2.5
0.59 (P < 0.01)
Ischemic validation cohort
ECV-guided LGE analysis versus FWHM
10
10 (100) versus 10 (100)
25.2 ± 8.5 versus 23.5 ± 5.7
25.0 (17.3–33.9) versus 24.6 (18.2–27.7) (P = 0.23)
1.8 ± 7.8
0.82 (P < 0.01)
Inter-method agreement was investigated in 80 cases of non-ischemic training cohort, 20 cases of non-ischemic validation cohort, and in 10 cases of ischemic validation cohort. A moderate correlation of scar amount (%) was observed between the ECV-guided LGE analysis and the manual analysis in the non-ischemic training cohort as well as in the validation cohort. In ischemic cases, the correlation was excellent
ECV extracellular volume, LGE late gadolinium enhancement, SD standard deviation, IQR interquartile range, LoA limits of agreement, CCC concordance correlation coefficient, FWHM full-width half-maximum
