AI-Based RGB Image Generation From RG Fundus Images Using Pix2Pix: Validation by Quantitative and Observer-Based Evaluations
Translational Vision Science & Technology · 2026
DOI: 10.1167/tvst.15.1.14 ↗Kurz gefasst
Validierungsstudie in der Zeitschrift Translational Vision Science & Technology, 2026. Verfasst von Kumiko Kato und 9 weiteren Autor:innen. Beteiligte Einrichtung: Department of Ophthalmology, Mie University Graduate School of Medicine, Tsu, Japan. Hauptthemen laut MEDLINE: Fundus Oculi, Image Processing, Computer-Assisted, Artificial Intelligence. Die Arbeit nennt 16 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Purpose
The purpose of this study was to evaluate the accuracy and perceptual quality of red-green-blue (RGB) fundus images generated from red-green (RG) images using Pix2Pix-based generative adversarial network (GAN).
Methods
RG images were extracted from original RGB images recorded by a newer model of the Optos system by turning off the blue laser source. A conditional GAN based on the Pix2Pix architecture was trained to convert this extracted RG image to an RGB image. The performance was evaluated quantitatively by using a five-fold cross-validation and assessments with structural similarity index measure (SSIM), peak signal-to-noise ratio (PSNR), learned perceptual image patch similarity (LPIPS), mean absolute error (MAE), and root mean square error (RMSE). To assess the perceptual accuracy, 39 ophthalmologists examined 60 randomly selected images, 30 true and 30 generated RGB images, using Google Form-based interface. They were instructed to classify each image as a true or an artificial intelligence (AI)-generated image.
Results
The generated images had high similarity to the original RGB images (SSIM = 0.97 ± 0.00; PSNR = 35.74 ± 0.17 decibel [dB]; LPIPS = 0.09 ± 0.00; MAE = 1.76 ± 0.04; and RMSE = 4.25 ± 0.10). The overall correct classification rate was 57.1%. Certain lesion types, such as epiretinal membrane and optic disc cupping, had higher discrimination rates. The receiver operating characteristic (ROC) analysis across all evaluations yielded an area under the curve (AUC) of 0.497 (P = 0.314).
Conclusions
Pix2Pix-based GANs can generate perceptually and structurally consistent RGB images from RG images without lesion-specific attention mechanisms. Translational Relevance: This method enables a reinterpretation of legacy RG fundus images and may support future AI diagnostic applications.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Ophthalmology, Mie University Graduate School of Medicine, Tsu, Japan.
- https://orcid.org/0000-0003-4172-3443.
- Department of Electrical and Electronic Engineering, Mie University, Tsu, Japan.
- Department of Ophthalmology, Keio University School of Medicine, Tokyo, Japan.
Zitieren
APA
Kato, K., Imai, K., Matsui, Y., Kitajima, T., Hirano, H., Mizumoto, K., Mase, Y., Irie-Ota, A., Kawanaka, H., & Kondo, M. (2026). AI-Based RGB Image Generation From RG Fundus Images Using Pix2Pix: Validation by Quantitative and Observer-Based Evaluations. Translational Vision Science & Technology. https://doi.org/10.1167/tvst.15.1.14
Vancouver
Kato K, Imai K, Matsui Y, Kitajima T, Hirano H, Mizumoto K, et al. AI-Based RGB Image Generation From RG Fundus Images Using Pix2Pix: Validation by Quantitative and Observer-Based Evaluations. Translational Vision Science & Technology. 2026. doi:10.1167/tvst.15.1.14
BibTeX
@article{Kato2026,
author = {Kato, Kumiko and Imai, Koki and Matsui, Yoshitsugu and Kitajima, Takumi and Hirano, Haru and Mizumoto, Keitaro and Mase, Yoko and Irie-Ota, Akiko and Kawanaka, Hiroharu and Kondo, Mineo},
title = {AI-Based RGB Image Generation From RG Fundus Images Using Pix2Pix: Validation by Quantitative and Observer-Based Evaluations},
journal = {Translational Vision Science & Technology},
year = {2026},
doi = {10.1167/tvst.15.1.14}
}
EUVIMED