Skip to main content
. 2021 Oct 21;8:732893. doi: 10.3389/fcvm.2021.732893

Table 2.

Comparison between cardiac magnetic resonance (CMR) and three-dimensional (3D) auto right ventricular (RV) quantification with or without manual editing.

Variable Novel software (n = 149) CMR (n = 149) Correlation p-Value Bias LOA
RVEF (%)
3D auto RV 38.9 (27.6–50.1) 34.0 (17.5–44.5) r = 0.781 <0.0001 6.8 19.2 (−12.4 to 26.0)
3D auto edit 35.6 (22.9–45.6) 34.0 (17.5–44.5) r = 0.941 <0.0001 2.6 10.2 (−7.6 to 12.8)
RVEDV (ml)
3D auto RV 112.9 (84.6–150.0) 119.8 (91.1–175.8) r = 0.794 <0.0001 −17.8 94.8 (−112.6 to 77.0)
3D auto edit 116.9 (88.6–148.9) 119.8 (91.1–175.8) r = 0.924 <0.0001 −12.3 66.8 (−79.1 to 54.5)
RVESV (ml)
3D auto RV 64.7 (42.9–110.3) 78.1 (51.7–147.7) r = 0.854 <0.0001 −23.6 93.6 (−117.2 to 70.0)
3D auto edit 73.6 (48.1–113.7) 78.1 (51.7–147.7) r = 0.955 <0.0001 −13.8 59.9 (−73.7 to 46.1)

Data are expressed as median and interquartile range. LOA, limit of agreement.