Objectives
Methods
Patient population
CMR image acquisition and analysis
Statistical methods
Results
Population characteristics
n = 125 | |
---|---|
Traditional CMR features | |
LV ejection fraction (%) | 52 ± 14 |
LV ejection fraction < 50% | 42 (34%) |
RV ejection fraction (%) | 51 ± 13 |
RV ejection fraction < 50% | 45 (37%) |
Pericardial effusion | 33 (26%) |
Late gadolinium enhancement | 65 (52%) |
Extent (%) | 3.9 ± 6.7 |
Elevated T2-ratio (> 2.0) | 17/67 (25%) |
Imaging features | |
GE scanner | 51 (41%) |
Siemens scanner | 74 (59%) |
1.5 Tesla | 68 (54%) |
3 Tesla | 57 (46%) |
≥ 30 phases | 48 (38%) |
Number of SAX slices per patient | 8 [7-9] |
Number of LAX slices per patient | 3 [2-3] |
Poor gating | 16 (13%) |
Time between ciné series (min) | 23 [9-33] |
Ciné images acquired: | |
• Pre-contrast • Mixed pre- and post-contrast • Post-contrast | 31 (25%) 27 (22%) 67 (54%) |
Inclusion of CMR-FT data
Reader Reliability and Agreement
Inter-reader | Intra-reader | ||||||
---|---|---|---|---|---|---|---|
Mean + SD | ICC (95%CI) | Absolute Agreement |Δ| | Relative Agreement % | ICC (95%CI) | Absolute Agreement |Δ| | Relative Agreement % | |
Global peak strain (%) | |||||||
2D-LV longitudinal | − 13.2 ± 4.0 | 0.96 (0.91–0.98)* | 1.3 ± 1.1 | 10 ± 9 | 0.96 (0.94–0.97)* | 1.1 ± 1.2 | 9 ± 8 |
3D-LV longitudinal | − 10.9 ± 4.0 | 0.93 (0.89–0.95)* | 1.6 ± 1.5 | 17 ± 26 | 0.93 (0.90–0.96)* | 1.5 ± 1.4 | 18 ± 27 |
RV longitudinal | − 18.5 ± 5.4 | 0.83 (0.75–0.90)* | 2.7 ± 3.0 | 16 ± 19 | 0.91 (0.87–0.94)* | 2.1 ± 2.4 | 13 ± 17 |
2D-LV circumferential | − 14.7 ± 4.7 | 0.99 (0.98–0.99)* | 0.8 ± 0.7 | 6 ± 6 | 0.99 (0.99–0.99)* | 0.6 ± 0.6 | 4 ± 5 |
3D-LV circumferential | − 16.2 ± 5.1 | 0.98 (0.96–0.99)* | 1.2 ± 1.1 | 8 ± 12 | 0.98 (0.97–0.99)* | 1.1 ± 1.1 | 7 ± 8 |
Early diastolic strain rate (/s) | |||||||
2D-LV longitudinal | 0.71 ± 0.25 | 0.86 (0.79–0.90)* | 0.13 ± 0.13 | 20 ± 22 | 0.92 (0.89–0.95)* | 0.11 ± 0.10 | 16 ± 16 |
3D-LV longitudinal | 0.65 ± 0.26 | 0.71 (0.57–0.80)* | 0.16 ± 0.20 | 24 ± 30 | 0.77 (0.67–0.84)* | 0.17 ± 0.19 | 26 ± 25 |
RV longitudinal | 1.40 ± 0.90 | 0.86 (0.80–0.91)* | 0.54 ± 0.87 | 25 ± 27 | 0.85 (0.77–0.90)* | 0.50 ± 0.73 | 21 ± 19 |
2D-LV circumferential | 0.84 ± 0.33 | 0.94 (0.91–0.96)* | 0.08 ± 0.14 | 10 ± 16 | 0.95 (0.93–0.97)* | 0.08 ± 0.13 | 10 ± 16 |
3D-LV circumferential | 0.97 ± 0.38 | 0.93 (0.90–0.95)* | 0.12 ± 0.16 | 14 ± 21 | 0.93 (0.90–0.95)* | 0.13 ± 0.16 | 15 ± 21 |
Factors impacting agreement
Inter-reader | Intra-reader | |||||||
---|---|---|---|---|---|---|---|---|
Univariable | Multivariable | Univariable | Multivariable | |||||
β | p | β | p | β | p | β | p | |
Peak strain | ||||||||
Δ2D LV | ||||||||
Baseline 2D-GLS | − 0.71 | 0.027 | − 1.09 | < 0.001 | − 0.09 | < 0.001 | ||
LV ejection fraction | 3.01 | 0.009 | 0.01 | 0.047 | 3.20 | 0.003 | ||
End diastolic volume | − 12.4 | 0.019 | − 18.1 | < 0.001 | ||||
Slice number | − 0.10 | 0.037 | − 0.13 | 0.004 | − 0.43 | 0.028 | ||
LGE (%) | − 1.17 | 0.033 | ||||||
Phases (≥ 30) | − 0.47 | 0.044 | ||||||
Temporal resolution* | 1.25 | 0.073 | ||||||
Δ3D LV | ||||||||
Siemens (vs. GE) | − 0.06 | 0.035 | − 0.06 | 0.091 | ||||
Heartrate | − 2.87 | 0.004 | ||||||
Slice number | − 0.06 | 0.058 | − 0.49 | 0.035 | − 0.11 | 0.008 | − 0.46 | 0.030 |
LGE (%) | − 0.68 | 0.086 | − 0.87 | 0.057 | ||||
End diastolic volume | − 9.42 | 0.014 | − 8.65 | 0.052 | ||||
LV ejection fraction | 2.19 | 0.010 | 2.14 | 0.032 | ||||
Phases (≥ 30) | 0.065 | 0.026 | 0.08 | 0.026 | ||||
Temporal resolution* | 1.78 | < 0.001 | ||||||
Year | − 0.60 | 0.030 | ||||||
Poor 3D reconstruction | 0.02 | 0.063 | ||||||
ΔRV | ||||||||
3 Tesla | 0.05 | 0.048 | ||||||
Gating (arrhythmia) | 0.02 | 0.037 | 2.12 | 0.030 | 0.03 | 0.010 | 2.02 | 0.008 |
Temporal resolution* | 0.781 | 0.040 | 0.05 | 0.089 | ||||
RV stroke volume | 1.98 | 0.094 | ||||||
Early diastolic strain rate | ||||||||
Δ2D LV | ||||||||
Heartrate | 26.0 | 0.033 | 30.9 | 0.043 | 0.001 | 0.015 | ||
Temporal resolution* | − 12.9 | 0.048 | − 14.3 | 0.082 | ||||
Slice number | − 0.95 | 0.012 | − 0.02 | 0.015 | − 1.12 | 0.033 | − 0.04 | 0.019 |
Stroke volume | − 39.1 | 0.043 | − 0.002 | 0.010 | ||||
LGE extent (%) | − 9.76 | 0.048 | − 0.003 | 0.073 | − 13.2 | 0.029 | − 0.003 | 0.057 |
LGE presence | − 0.66 | 0.066 | ||||||
Baseline 2D-edSR | 0.296 | 0.091 | 0.62 | 0.005 | ||||
Siemens (vs GE) | − 0.88 | 0.045 | ||||||
Post contrast agent | 0.66 | 0.081 | 0.55 | 0.012 | ||||
Phases (≥ 30) | 0.84 | 0.056 | ||||||
Δ3D LV | ||||||||
Poor 3D construction | 0.20 | 0.014 | 0.32 | 0.005 | ||||
Pericardial effusion | 0.40 | 0.066 | 0.08 | 0.042 | 0.61 | 0.005 | 0.11 | 0.002 |
LGE presence | − 0.49 | 0.055 | – | – | – | – | ||
Baseline 3D-edSR | 0.36 | 0.003 | 0.25 | 0.002 | 0.57 | < 0.001 | 0.31 | < 0.001 |
ΔRV | ||||||||
3-Tesla | 1.19 | 0.002 | 0.10 | 0.016 | ||||
Siemens (vs GE) | 0.46 | 0.083 | ||||||
Post contrast agent | 0.461 | 0.051 | ||||||
Phases (≥ 30) | − 0.52 | 0.053 | ||||||
Pericardial effusion | 0.48 | 0.048 | ||||||
Heartrate | 23.9 | 0.005 | 0.003 | 0.024 | ||||
Temporal resolution* | –9.16 | 0.044 | ||||||
Baseline RV-edSR | 0.47 | 0.005 | 0.14 | 0.021 |
Factor | Inter-reader | Intra-reader | ||||||
---|---|---|---|---|---|---|---|---|
Univariable | Multivariable | Univariable | Multivariable | |||||
β | p | β | p | β | p | β | p | |
Peak strain | ||||||||
Δ2D LV | ||||||||
LV ejection fraction | 4.47 | 0.010 | 0.01 | 0.014 | ||||
Stroke volume | 5.53 | 0.090 | ||||||
Myocardial mass | 11.9 | 0.017 | 0.007 | < 0.001 | ||||
LGE presence | –0.14 | 0.020 | 0.35 | 0.009 | ||||
Gating (arrhythmia) | 0.09 | 0.034 | 0.43 | 0.023 | ||||
Slice number | –0.46 | 0.008 | –0.12 | 0.005 | –0.56 | 0.012 | –0.07 | 0.048 |
Year | –1.38 | 0.028 | ||||||
Siemens (vs GE) | –0.14 | 0.078 | ||||||
Phases (≥ 30) | 00.17 | 0.026 | ||||||
Δ3D LV | ||||||||
Slice number | –0.38 | 0.001 | –0.15 | 0.064 | –0.29 | 0.015 | –0.17 | 0.016 |
End diastolic volume | –11.4 | 0.031 | ||||||
LV ejection fraction | 1.93 | 0.100 | ||||||
Poor 3D-reconstruction | 0.03 | 0.028 | 0.99 | 0.085 | ||||
Temporal resolution* | 1.34 | 0.077 | 0.02 | 0.052 | ||||
Early diastolic strain rate | ||||||||
Δ2D LV | ||||||||
Heart rate | 20.9 | 0.080 | 0.001 | 0.080 | ||||
LGE presence | –0.58 | 0.075 | 0.05 | 0.075 | ||||
Δ3D LV | ||||||||
Year | 7.0 | 0.002 | ||||||
Siemens (vs GE) | 0.49 | 0.086 | ||||||
Post contrast agent | –0.43 | 0.068 | ||||||
Phases (≥ 30) | –0.52 | 0.069 | ||||||
Time between images | –14.8 | 0.068 | ||||||
Pericardial effusion | 0.59 | 0.020 | 0.11 | 0.048 |