Ejection Fraction Prediction on EchoNet-Dynamic (test)
0.84R2MCSDR
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MCSDRType=Generative Regression, Variant=Vision only2026.02 | 0.84 | 3.76 | 4.81 | |
| R(2+1)D2023.10 | 0.822 | — | — | |
| EchoCoTr-SNo. of Frames=36, FLOPs=19.611G, sampling frequency=42022.09 | 0.82 | 3.95 | 5.17 | |
| CoReEchoClip(s)=3 x Clips2024.03 | 0.82 | 3.9 | 5.13 | |
| EchoCoTrType=Deterministic Regression2026.02 | 0.82 | 3.95 | 5.17 | |
| CoReEchoType=Deterministic Regression2026.02 | 0.82 | 3.9 | 5.13 | |
| FHFA-MTLType=Deterministic Regression2026.02 | 0.82 | 3.89 | 5.13 | |
| EchoGraphs with regression headInput=single heartbeat, Frames=162022.07 | 0.81 | 4.01 | 5.36 | |
| EchoNetInput=whole video2022.07 | 0.81 | 4.05 | 5.32 | |
| EchoNet-Dynamic (2)No. of Frames=beat-to-beat, Evaluation protocol=beat-to-beat2022.09 | 0.81 | 4.05 | 5.32 | |
| EchoCoTr-BNo. of Frames=36, FLOPs=44.907G, sampling frequency=42022.09 | 0.81 | 3.98 | 5.34 | |
| Claimedsource=original reported results, trained_on=real data2024.06 | 0.81 | 4.05 | 5.32 | |
| Reproducedsource=locally reproduced, trained_on=real data2024.06 | 0.81 | 3.98 | 5.29 | |
| [19]Clip(s)=Full video2024.03 | 0.81 | 4.05 | 5.32 | |
| [26]Clip(s)=1 x SH2024.03 | 0.81 | 4.01 | 5.36 | |
| [3]Clip(s)=3 x Clips, reproduced=true2024.03 | 0.81 | 4.03 | 5.28 | |
| EchoCoTrClip(s)=3 x Clips, reproduced=true2024.03 | 0.81 | 3.98 | 5.3 | |
| EchoGraphsType=Deterministic Regression2026.02 | 0.81 | 4.01 | 5.36 | |
| VAE rec.trained_on=VAE-reconstructed real data2024.06 | 0.8 | 4.12 | 5.48 | |
| EchoNet (MC3)Input=single heartbeat, Frames=322022.07 | 0.79 | 4.22 | 5.56 | |
| EchoGraphs - ED/ES classifierInput=whole video2022.07 | 0.79 | 4.23 | 5.67 | |
| R3DNo. of Frames=32, FLOPs=92.273G2022.09 | 0.79 | 4.22 | 5.62 | |
| EchoNet-Dynamic (1)No. of Frames=32, FLOPs=91.974G, Evaluation protocol=clip-level2022.09 | 0.79 | 4.22 | 5.56 | |
| [19]Clip(s)=1 x SH2024.03 | 0.79 | 4.22 | 5.56 | |
| [26]Clip(s)=Full video2024.03 | 0.79 | 4.23 | 5.67 | |
| EchoNet(32F)Type=Deterministic Regression2026.02 | 0.79 | 4.22 | 5.56 | |
| MC3No. of Frames=32, FLOPs=97.656G2022.09 | 0.77 | 4.54 | 5.97 | |
| USVNNa=1282023.10 | 0.765 | — | — | |
| EchoGNNType=Deterministic Regression2026.02 | 0.76 | 4.45 | 5.97 | |
| Ourstrained_on=our LVDM generated synthetic data, tested_on=real data2024.06 | 0.75 | 4.55 | 6.1 | |
| EchoGraphs using keypointsInput=single heartbeat, Frames=162022.07 | 0.73 | 4.66 | 6.3 | |
| Regression onlyInput=single heartbeat, Frames=162022.07 | 0.72 | 4.28 | 5.75 | |
| Average Pool2023.10 | 0.679 | — | — | |
| Max Pool2023.10 | 0.657 | — | — | |
| EchoGraphs - Peak computationInput=whole video2022.07 | 0.65 | 4.3 | 5.86 | |
| Transformer (M.)Input=single heartbeat, Frames=1282022.07 | 0.64 | 5.32 | 7.23 | |
| [20]Clip(s)=1 x SH2024.03 | 0.64 | 5.32 | 7.23 | |
| Batool et al.Type=Deterministic Regression2026.02 | 0.6 | 5.74 | 7.73 | |
| EchoDiff. [24]trained_on=synthetic data from [24]2024.06 | 0.55 | 6.02 | 8.21 | |
| TransformerInput=whole video2022.07 | 0.52 | 5.95 | 8.38 | |
| UVT M.No. of Frames=128, FLOPs=130.00G, Sampling strategy=Mirroring2022.09 | 0.52 | 5.95 | 8.38 | |
| [20]Clip(s)=Full video2024.03 | 0.52 | 5.95 | 8.38 | |
| UVTType=Deterministic Regression2026.02 | 0.52 | 5.95 | 8.38 | |
| UVT R.No. of Frames=128, FLOPs=130.00G, Sampling strategy=Random2022.09 | 0.48 | 6.77 | 8.7 | |
| EchoClipType=Foundation Model2026.02 | — | 7.14 | — | |
| PanEchoType=Foundation Model2026.02 | — | 5.35 | — |