Table 6.

Summary of literature review for post-intervention prediction task using GAN in ophthalmology imaging domains

Publication	Basic technique	Domain	Intervention	Summary
Yoo et al. [15]	Conditional GAN, CycleGAN	Periorbital facial images	Orbital decompression surgery	The developed model transformed preoperative facial input images into predicted postoperative images for orbital decompression for thyroid-associated ophthalmopathy
Liu et al. [66]	Pix2pix (conditional GAN)	Retinal OCT	Intravitreal anti-vascular endothelial growth factor injection	The model generated individualized post-therapeutic OCT images that could predict the short-term response of treatment for age-related macular degeneration
Lee et al. [67]	Conditional GAN (multi-channel inputs)	Retinal OCT (with fluorescein angiography and indocyanine green angiography)	Intravitreal anti-vascular endothelial growth factor injection	The trained model generated post-treatment optical coherence tomography (OCT) images of neovascular age-related macular degeneration

GAN = generative adversarial network; OCT = optical coherence tomography