Table 6 Performance comparison for different values of \(N\) (number of fusion paths) in the multi-path fusion strategy. Values in bold indicate the best performance.
From: A novel multimodel medical image fusion framework with edge enhancement and cross-scale transformer
Fusion tasks | CT and MRI image fusion | PET and MRI image fusion | SPECT and MRI image fusion | ||||||
|---|---|---|---|---|---|---|---|---|---|
\(N=1\) | \(N=3\) | \(N=4\) | \(N=1\) | \(N=3\) | \(N=4\) | \(N=1\) | \(N=3\) | \(N=4\) | |
MI | 3.2559 | 3.6244 | 3.3282 | 3.9962 | 4.5539 | 4.5547 | 4.0444 | 4.5115 | 4.4145 |
Qabf | 0.5679 | 0.6497 | 0.6406 | 0.7501 | 0.7555 | 0.7558 | 0.7346 | 0.7378 | 0.7381 |
VIF | 0.5671 | 0.6707 | 0.6109 | 0.8233 | 0.8788 | 0.8834 | 0.8545 | 0.8796 | 0.8825 |
SSIM | 0.6451 | 0.6554 | 0.6703 | 0.6375 | 0.6487 | 0.6483 | 0.6135 | 0.6152 | 0.6107 |
Qcb | 0.6804 | 0.6981 | 0.6916 | 0.6230 | 0.6552 | 0.6535 | 0.6788 | 0.6779 | 0.6744 |
Qcv\(\downarrow\) | 1288.90 | 1150.07 | 1389.13 | 72.10 | 71.50 | 71.44 | 51.23 | 50.37 | 50.15 |
