Table 10 Comparison of the performance of GAN proposed in this study with state-of-the-art GANs using metrics of reference-based category.

From: A generative adversarial network for synthetization of regions of interest based on digital mammograms

Authors and references

GAN model

SSIM

DSSIM

PSNR

MSE

69

Conditional GAN (cGAN)

0.8960

0.05

23.65

313.2

94

Peceptual GAN

0.9071

0.05

24.20

287

56

Style-content (SC-GAN)

0.9046

0.05

24.12

282.8

79

MedGAN

0.9160

0.04

24.62

264.8

This study

ROImammoGAN

0.8000

0.10

27.72

109.92

Back to article page