Table 2.
CMR-FT strain, conventional and AI-derived automated volume transient metrics.
| Variable | ALL patients | MACE (n = 73) | No MACE (n = 948) | AUCk | Pval | HR | HR Pval |
|---|---|---|---|---|---|---|---|
| Strain–Strain markers | |||||||
| Global longitudinal strain (%) | − 16.5 (− 20.2 to− 12.5) | − 11.5 (− 16.8–-8.4) | − 16.7 (− 20.4 to − 13.0) | 0.710 | < 0.001 | 2.13 (1.68–2.69) | < 0.001 |
| Global circumferential strain (%) | − 24.2 (− 28.9 to − 19.2) | − 18.6 (− 24.2–-14.6) | − 24.4 (− 29.1 to − 19.7) | 0.674 | < 0.001 | 1.72 (1.39–2.12) | < 0.001 |
| Global radial strain (%) | 20.4 (15.7–25.9) | 16.4 (12.6–22.6) | 20.5 (16.0–26.0) | 0.626 | < 0.001 | 0.64 (0.50–0.83) | < 0.001 |
| CURE | 0.78 (0.71–0.84) | 0.71 (00.7–0.80) | 0.79 (0.72–0.84) | 0.637 | < 0.001 | 0.62 (0.50–0.76) | < 0.001 |
| RURE | 0.75 (0.67–0.82) | 0.69 (00.6–0.78) | 0.76 (0.67–0.83) | 0.629 | < 0.001 | 0.66 (0.54–0.82) | < 0.001 |
| Vt—Volume transient markers | |||||||
| RR-interval (ms) | 845 (741–952) | 780 (674–848) | 845 (750–952) | 0.644 | < 0.001 | 0.64 (0.51–0.81) | < 0.001 |
| RR-interval variability (%RR) | 19.2 (13.1–26.3) | 24.6 (14.2–36.7) | 18.9 (13.0–25.6) | 0.606 | < 0.001 | 1.28 (1.13–1.46) | < 0.001 |
| Systolic phase | |||||||
| Systolic time (ms) | 297 (267–328) | 284 (265–321) | 298 (268–328) | 0.562 | 0.106 | 0.70 (0.51–0.98) | 0.038 |
| Systolic velocity avg (mL/s) | 251 (209–301) | 227 (184–271) | 252 (212–303) | 0.592 | 0.238 | 0.61 (0.34–1.11) | 0.105 |
| Systolic velocity avg (%SV/s) | 337 (305–375) | 352 (312–377) | 336 (305–373) | 0.554 | 0.314 | 1.08 (0.93–1.25) | 0.315 |
| Systolic velocity peak (mL/s) | 570 (451–704) | 514 (396–650) | 573 (456–711) | 0.573 | 0.030 | 0.77 (0.60–0.97) | 0.029 |
| Diastolic phase | |||||||
| Diastolic time (ms) | 513 (431–602) | 462 (394–523) | 518 (435–606) | 0.611 | 0.028 | 0.76 (0.60–0.97) | 0.025 |
| Diastasis (ms) | 38 (0–118) | 24 (0–83) | 44 (0–119) | 0.569 | 0.034 | 0.74 (0.56–0.98) | 0.037 |
| Diastasis (%) | 10.8 (0.0–21.6) | 5.4 (0.0–16.2) | 10.8 (0.0–21.6) | 0.558 | 0.030 | 0.74 (0.56–0.97) | 0.032 |
| Time to diastasis (ms) | 354 (307–402) | 327 (267–366) | 356 (309–405) | 0.605 | 0.017 | 0.70 (0.53–0.93) | 0.013 |
| Time to diastasis (%) | 64.9 (62.2–70.3) | 64.9 (56.8–70.9) | 64.9 (62.2–70.3) | 0.512 | 0.192 | 0.82 (0.62–1.09) | 0.177 |
| Passive filling (mL) | 55.6 (43.0–67.2) | 48.1 (37.6–60.1) | 56.0 (43.5–67.9) | 0.630 | 0.002 | 0.64 (0.48–0.85) | 0.002 |
| Passive filling (%SV) | 66.5 (58.2–73.5) | 67.3 (57.8–78.9) | 66.4 (58.2–73.0) | < 0.5 | 0.767 | 1.05 (0.77–1.42) | 0.776 |
| Active filling (mL) | 26.2 (19.9–33.6) | 20.6 (12.7–28.5) | 26.7 (20.2–33.9) | 0.647 | < 0.001 | 0.71 (0.59–0.87) | < 0.001 |
| Active filling (%SV) | 31.5 (23.8–39.8) | 26.6 (17.8–37.7) | 31.7 (24.1–39.9) | 0.559 | 0.110 | 0.81 (0.63–1.04) | 0.094 |
| Diastolic velocity avg (mL/s) | 143 (121–170) | 141 (109–168) | 144 (122–170) | 0.535 | 0.458 | 0.20 (0.04–0.97) | 0.046 |
| Diastolic velocity avg (%SV/s) | 195 (166–232) | 216 (191–254) | 193 (165–230) | < 0.5 | 0.982 | 1.00 (0.79–1.26) | 0.982 |
| Passive filling velocity avg (mL/s) | 155 (122–187) | 151 (118–180) | 155 (122–188) | 0.516 | 0.291 | 0.87 (0.69–1.11) | 0.274 |
| Passive filling velocity avg (%SV/s) | 185 (153–215) | 205 (170–239) | 184 (153–213) | 0.592 | 0.015 | 1.35 (1.07–1.71) | 0.013 |
| Passive filling velocity peak (mL/s) | 427 (307–573) | 384 (271–502) | 429 (310–575) | 0.553 | 0.143 | 0.80 (0.59–1.08) | 0.143 |
| Active filling velocity avg (mL/s) | 128 (92–169) | 104 (70–144) | 130 (94–169) | 0.610 | 0.011 | 0.76 (0.62–0.93) | 0.008 |
| Active filling velocity avg (%SV/s) | 154 (106–205) | 142 (94–222) | 154 (107–204) | 0.521 | 0.277 | 0.85 (0.65–1.12) | 0.260 |
| Active filling velocity peak (mL/s) | 315 (222–456) | 246 (176–382) | 323 (227–464) | 0.629 | 0.003 | 0.64 (0.47–0.85) | 0.003 |
| VtAI—Volume Transient modes: | |||||||
| VtAI1 | 0.03 (− 0.45–0.54) | 0.39 (− 00.2–0.66) | 0.02 (− 0.47–0.51) | 0.570 | 0.112 | 1.22 (0.96–1.56) | 0.103 |
| VtAI2 | − 0.08 (− 0.17–0.04) | 0.00 (-00.1–0.13) | − 0.08 (-0.18–0.04) | 0.598 | < 0.001 | 1.27 (1.12–1.44) | < 0.001 |
| VtAI3 | 0.00 (− 0.22–0.22) | 0.11 (-00.1–0.33) | − 0.01 (-0.22–0.21) | 0.588 | 0.025 | 1.34 (1.05–1.71) | 0.020 |
| VtAI5 | − 0.02 (− 0.10–0.09) | 0.06 (− 00.1–0.17) | − 0.02 (− 0.10–0.08) | 0.611 | < 0.001 | 1.35 (1.14–1.59) | < 0.001 |
Data presented as median (interquartile range). P-values were calculated for the comparison between patients with and without MACE. Univariate Cox regression analyses are presented as HR, Hazard ratios (95% confidence interval), and HR P-value, predictor significance. The predictive power of each biomarker is assessed via LDA and presented as median AUC 10-cross-fold validated, repeated for a hundred random data splits, AUCk. MACE indicates major adverse cardiac events; and avg, average. The passive and active filling metrics were not reported in those cases where diastasis was not found (diastasis = 0 s, n = 212).