Optics and Precision Engineering, Volume. 32, Issue 2, 221(2024)
PET/CT Cross-modal medical image fusion of lung tumors based on DCIF-GAN
Figures & Tables(12)
Fig. 1. Dual-coupled interactive fusion GAN overall network architecture
Table 1. Comparison experiment 1 fusion image index evaluation results
View tableView in ArticleTable 1. Comparison experiment 1 fusion image index evaluation results
图像组 方法 AG SF SSIM SD MI PSNR IE QAB/F 1 Fusion GAN 2.003 6.97 0.497 26.8 1.482 10.032 4.875 2 0.109 3 DDcGAN 7.477 21.32 0.587 33.11 1.432 10.251 6.663 5 0.347 6 MGMDcGAN 5.184 14.88 0.601 29.88 2.145 10.338 6.258 5 0.285 3 LatLRR+NSCT 8.071 23.27 0.600 45.48 1.512 10.331 7.328 7 0.548 2 OURS 8.706 23.37 0.608 59.04 1.897 10.422 7.395 0 0.497 3 2 Fusion GAN 1.78 6.2 0.535 28.24 1.354 10.133 5.120 0 0.105 9 DDcGAN 6.866 20.08 0.622 35.08 1.441 10.622 6.625 3 0.338 0 MGMDcGAN 4.748 14.34 0.630 31.53 1.95 10.657 6.163 2 0.276 4 LatLRR+NSCT 7.320 0 22.020 0 0.531 46.03 1.386 10.644 7.306 2 0.531 8 OURS 8.113 22.3 0.698 62.81 1.861 11.363 7.482 4 0.486 8 3 Fusion GAN 2.893 9.86 0.429 37.08 1.73 8.839 5.681 1 0.127 4 DDcGAN 7.087 21.74 0.499 35.56 2.514 8.906 6.155 3 0.336 3 MGMDcGAN 4.985 15.77 0.492 32.03 2.925 9.006 5.830 8 0.285 0 LatLRR+NSCT 8.305 23.4 0.492 46.68 2.104 8.940 7.113 0 0.559 6 OURS 8.388 23.99 0.574 66.75 2.029 9.740 7.502 8 0.526 4 4 Fusion GAN 1.893 6.29 0.517 25.13 1.428 10.286 5.120 8 0.096 6 DDcGAN 7.84 22.12 0.590 36.61 1.711 10.322 6.886 5 0.338 3 MGMDcGAN 5.565 15.63 0.601 32.84 2.071 10.419 6.529 6 0.291 1 LatLRR+NSCT 8.668 23.43 0.604 50.14 1.548 10.425 7.474 5 0.563 2 OURS 9.071 23.94 0.666 62.78 1.828 10.515 7.638 3 0.496 7 5 Fusion GAN 2.034 6.69 0.500 29.19 1.324 9.577 5.262 4 0.084 0 DDcGAN 8.019 21.69 0.558 36.91 1.994 9.778 6.996 9 0.343 5 MGMDcGAN 5.534 15.19 0.571 33.21 2.318 9.882 6.607 6 0.274 9 LatLRR+NSCT 8.76 24.33 0.573 50.23 1.747 9.861 7.547 2 0.583 4 OURS 9.329 24.61 0.580 63.85 1.923 9.949 7.567 9 0.506 4
Table 2. Comparison experiment 2 fusion image index evaluation results
View tableView in ArticleTable 2. Comparison experiment 2 fusion image index evaluation results
图像组 方法 AG SF SSIM SD MI PSNR IE QAB/F 1 Fusion GAN 4.877 14.56 0.303 32.42 1.736 8.234 5.982 8 0.120 7 DDcGAN 10.897 25.58 0.381 27.48 1.611 8.376 6.471 3 0.289 5 MGMDcGAN 7.440 17.34 0.384 23.63 0.757 1 8.489 6.232 1 0.253 1 LatLRR+NSCT 13.582 30.49 0.384 38.87 1.524 8.469 7.491 3 0.551 9 OURS 13.619 30.85 0.391 52.81 2.05 8.522 7.752 6 0.518 0 2 Fusion GAN 4.665 14.96 0.333 35.46 2.441 8.12 6.012 8 0.200 9 DDcGAN 7.522 22.76 0.403 30.27 2.322 8.155 6.185 3 0.314 2 MGMDcGAN 5.082 15.64 0.388 27.27 0.549 5 8.255 5.976 2 0.269 0 LatLRR+NSCT 8.235 24.45 0.407 40.27 2.124 8.173 7.049 1 0.550 3 OURS 8.322 24.55 0.407 52.68 2.914 8.266 7.387 5 0.488 8 3 Fusion GAN 4.215 13.08 0.279 31.19 1.811 7.714 5.469 3 0.120 2 DDcGAN 9.933 26.19 0.367 27.53 1.842 7.755 5.681 7 0.290 6 MGMDcGAN 6.909 18.32 0.36 22.9 0.703 4 7.859 5.457 9 0.245 5 LatLRR+NSCT 11.858 31.09 0.382 35.890 0 1.247 7.772 6.901 1 0.510 3 OURS 12.231 31.14 0.383 44.55 2.216 7.873 7.276 7 0.461 8 4 Fusion GAN 4.291 13.72 0.286 30.4 1.676 7.752 5.375 9 0.121 4 DDcGAN 9.783 25.47 0.374 26.24 1.656 7.852 5.626 8 0.276 MGMDcGAN 6.913 18.11 0.367 21.9 0.728 9 7.954 5.420 0 0.243 6 LatLRR+NSCT 12.234 31.78 0.394 34.82 1.154 7.864 0 6.925 4 0.512 1 OURS 12.42 31.25 0.396 46.6 2.006 7.961 7.291 5 0.468 8 5 Fusion GAN 5.015 16.28 0.317 36.79 2.217 7.495 5.473 6 0.164 7 DDcGAN 8.253 24.19 0.375 33.53 2.417 7.637 5.52 0.298 1 MGMDcGAN 5.372 16.15 0.366 29.67 0.610 87 7.737 5.322 2 0.239 7 LatLRR+NSCT 9.324 27.06 0.38 43.610 0 1.858 7.671 6.736 8 0.533 6 OURS 9.404 27.14 0.382 55.23 2.599 7.747 7.022 1 0.478 8
Table 3. Evaluation index of ablation experimental results.
View tableView in ArticleTable 3. Evaluation index of ablation experimental results.
图像组 方法 AG SF SSIM SD MI PSNR IE QAB/F 1 Network1 8.159 23.34 0.352 39.14 2.525 7.613 4.943 4 0.294 7 Network2 5.951 20.04 0.354 38.94 2.537 7.602 4.721 9 0.248 4 Network3 7.899 24.99 0.346 54.4 2.609 7.674 0 5.518 4 0.432 4 Network4 9.366 27.18 0.355 57.04 2.61 7.799 6.505 6 0.473 2 2 Network1 7.845 24.42 0.312 33.17 2.135 7.284 4.619 9 0.302 4 Network2 5.644 19.73 0.315 32.56 2.219 7.278 4.528 0 0.238 6 Network3 7.369 25.96 0.328 47.13 2.617 7.494 5.260 3 0.367 3 Network4 8.615 26.00 0.333 52.53 2.971 7.523 6.454 7 0.460 3 3 Network1 8.974 26.4 0.309 33.11 2.647 7.278 4.802 8 0.320 3 Network2 5.742 20.27 0.306 31.26 2.766 7.273 4.536 4 0.215 1 Network3 8.233 27.25 0.315 45.13 2.904 7.444 5.357 2 0.452 8 Network4 10.247 30.11 0.323 47.40 2.996 7.504 6.332 0 0.484 0 4 Network1 7.266 20.4 0.567 38.99 1.447 10.074 6.656 4 0.363 1 Network2 6.052 19.26 0.564 40.67 1.617 10.059 6.443 9 0.339 9 Network3 7.581 22.10 0.579 56.84 1.801 10.253 6.867 1 0.463 1 Network4 8.024 22.16 0.568 59.04 1.897 10.302 6.867 3 0.485 0 5 Network1 6.623 19.42 0.587 43.87 1.826 10.397 6.382 4 0.338 6 Network2 5.9 20.38 0.584 49.22 1.966 10.318 6.147 7 0.334 0 Network3 6.606 21.21 0.603 64.77 2.08 10.435 6.232 5 0.393 7 Network4 7.209 21.69 0.608 65.43 2.136 10.766 6.711 2 0.457 2
Tools
Get Citation
Copy Citation Text
Tao ZHOU, Qianru CHENG, Xiangxiang ZHANG, Qi LI, Huiling LU. PET/CT Cross-modal medical image fusion of lung tumors based on DCIF-GAN[J]. Optics and Precision Engineering, 2024, 32(2): 221
Paper Information
Category:
Received: Aug. 2, 2023
Accepted: --
Published Online: Apr. 2, 2024
The Author Email: CHENG Qianru (chengqianru5@163. com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20