Research Article
A Multifeature Extraction Method Using Deep Residual Network for MR Image Denoising
Table 1
SSIM comparison of different denoising algorithms.
| Dataset | Algorithm | 1% | 4% | 7% | 10% | 13% |
| TCGA-GBM | Ref. [14] | 0.9903 | 0.9584 | 0.9153 | 0.8612 | 0.8109 | Ref. [20] | 0.9901 | 0.9592 | 0.9276 | 0.8928 | 0.8529 | Ref. [26] | 0.9918 | 0.9643 | 0.9374 | 0.9022 | 0.8657 | The proposed algorithm | 0.9941 | 0.9668 | 0.9417 | 0.9152 | 0.8893 |
| CH-GBM | Ref. [14] | 0.9829 | 0.9622 | 0.9238 | 0.8874 | 0.8378 | Ref. [20] | 0.9923 | 0.9637 | 0.9246 | 0.8984 | 0.8557 | Ref. [26] | 0.9935 | 0.9626 | 0.9415 | 0.9043 | 0.8736 | The proposed algorithm | 0.9978 | 0.9689 | 0.9426 | 0.9037 | 0.8809 |
|
|