Table 2.
Results of objective rating. Abbreviations: CS = conventional reconstructions of Compressed SENSE images, CS AI = deep-learning-based reconstructions of Compressed SENSE images.
| Standard | CS 4.5 | CS AI 4.5 | CS 8 | CS AI 8 | CS 11 | CS AI 11 | |
|---|---|---|---|---|---|---|---|
| RMSE | - | 79.80 ± 37.03 | 80.53 ± 36.77 | 90.66 ± 41.37 | 91.08 ± 40.26 | 94.87 ± 44.07 | 94.79 ± 43.49 |
| SSIM | - | 0.86 ± 0.04 | 0.86 ± 0.04 | 0.84 ± 0.04 | 0.84 ± 0.04 | 0.83 ± 0.04 | 0.83 ± 0.04 |
| aSNR bone | 6.00 ± 1.69 | 6.51 ± 1.79 | 6.60 ± 1.97 | 5.60 ± 1.38 | 6.18 ± 1.64 | 5.05 ± 1.21 | 5.72 ± 1.70 |
| aSNR spinal cord | 11.36 ± 4.18 | 10.39 ± 3.65 | 10.19 ± 3.79 | 10.56 ± 3.71 | 10.64 ± 3.61 | 10.40 ± 3.36 | 10.22 ± 3.59 |
| aSNR CSF | 26.26 ± 14.05 | 28.17 ± 13.53 | 28.58 ± 13.58 | 26.89 ± 13.17 | 27.01 ± 13.06 | 26.42 ± 12.91 | 27.69 ± 13.79 |
| aCNR bone/CSF | 17.41 ± 7.88 | 19.70 ± 7.70 | 19.79 ± 7.72 | 18.46 ± 7.24 | 18.83 ± 7.25 | 17.63 ± 6.48 | 18.57 ± 6.60 |
| aCNR spinal cord/CSF | 21.72 ± 11.39 | 23.18 ± 9.90 | 23.36 ± 9.89 | 22.26 ± 9.79 | 22.27 ± 9.79 | 21.92 ± 9.58 | 22.63 ± 9.43 |