Accelerated MRI reconstruction on fastMRI knee multi-coil x8 (test)
89.51SSIMHUMUS-Net-L
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| HUMUS-Net-L# of params (approx.)=228M, Training data setup=fastMRI combined training and validation dataset2022.03 | 89.51 | 37.4 | 0.008 | |
| HUMUS-Net# of params (approx.)=109M, Training data setup=fastMRI combined training and validation dataset2022.03 | 89.45 | 37.3 | 0.0081 | |
| HUMUS-Net-L# of params (approx.)=228M, Training data setup=fastMRI training dataset2022.03 | 89.44 | 37.3 | 0.0083 | |
| HUMUS-Net# of params (approx.)=109M, Training data setup=fastMRI training dataset2022.03 | 89.36 | 37 | 0.0086 | |
| E2E-VarNet# of params (approx.)=30M, Training data setup=fastMRI combined training and validation dataset2022.03 | 89.2 | 37.1 | 0.0085 | |
| E2E-VarNet# of params (approx.)=30M, Training data setup=fastMRI training dataset2022.03 | 89 | 36.9 | 0.0089 | |
| XPDNet# of params (approx.)=155M, Training data setup=fastMRI training dataset2022.03 | 88.93 | 37.2 | 0.0083 | |
| Σ-Net# of params (approx.)=676M, Training data setup=fastMRI training dataset2022.03 | 88.77 | 36.7 | 0.0091 | |
| i-RIM# of params (approx.)=300M, Training data setup=fastMRI training dataset2022.03 | 88.75 | 36.7 | 0.0091 | |
| U-Net# of params (approx.)=214M, Training data setup=fastMRI training dataset2022.03 | 86.4 | 34.7 | 0.0132 |