Table 1 Quantitative comparison of reconstruction methods on the AAPM simulated dataset with 64 and 32 views. Best results are shown in bold.
From: Dual-Domain deep prior guided sparse-view CT reconstruction with multi-scale fusion attention
Method | FBP | FBPConvNet | FISTA | DUAL | FreeSeed | MVMS-RCN | DRONE | DMAprior | DPMA |
|---|---|---|---|---|---|---|---|---|---|
64 Views | |||||||||
PSNR | 26.135 | 35.623 | 39.121 | 40.231 | 42.348 | 42.516 | 43.027 | 41.526 | 43.867 |
SSIM | 0.534 | 0.933 | 0.965 | 0.971 | 0.977 | 0.980 | 0.982 | 0.975 | 0.984 |
32 Views | |||||||||
PSNR | 23.115 | 32.790 | 33.796 | 34.196 | 37.613 | 38.297 | 38.288 | 37.191 | 38.862 |
SSIM | 0.415 | 0.904 | 0.912 | 0.918 | 0.951 | 0.961 | 0.960 | 0.950 | 0.962 |
