Digital ray: enhancing cataractous fundus images using style transfer generative adversarial networks to improve retinopathy detection

被引：2

作者：

Liu, Lixue ^{[1
]}

Hong, Jiaming ^{[2
]}

Wu, Yuxuan ^{[1
]}

Liu, Shaopeng ^{[3
]}

Wang, Kai ^{[3
]}

Li, Mingyuan ^{[1
]}

Zhao, Lanqin ^{[1
]}

Liu, Zhenzhen ^{[1
]}

Li, Longhui ^{[1
]}

Cui, Tingxin ^{[1
]}

Tsui, Ching-Kit ^{[1
]}

Xu, Fabao ^{[4
]}

Hu, Weiling ^{[1
]}

Yun, Dongyuan ^{[1
]}

Chen, Xi ^{[1
]}

Shang, Yuanjun ^{[1
]}

Bi, Shaowei ^{[1
]}

Wei, Xiaoyue ^{[1
]}

Lai, Yunxi ^{[1
]}

Lin, Duoru ^{[1
]}

Fu, Zhe ^{[5
]}

Deng, Yaru ^{[5
]}

Cai, Kaimin ^{[5
]}

Xie, Yi ^{[5
]}

Cao, Zizheng ^{[1
]}

Wang, Dongni ^{[1
]}

Zhang, Xulin ^{[1
]}

Dongye, Meimei ^{[1
]}

Lin, Haotian ^{[1
]}

Wu, Xiaohang ^{[1
]}

机构：

[1] Sun Yat Sen Univ, Zhongshan Ophthalm Ctr, State Key Lab Ophthalmol, Guangzhou, Guangdong, Peoples R China

[2] Guangzhou Univ Chinese Med, Sch Med Informat Engn, Guangzhou, Guangdong, Peoples R China

[3] Guangdong Polytech Normal Univ, Sch Comp Sci, Guangzhou, Guangdong, Peoples R China

[4] Shandong Univ, Qilu Hosp, Jinan, Shandong, Peoples R China

[5] Sun Yat Sen Univ, Zhongshan Sch Med, Guangzhou, Guangdong, Peoples R China

来源：

BRITISH JOURNAL OF OPHTHALMOLOGY | 2024年 / 108卷 / 10期

关键词：

Lens and zonules; Imaging; Retina; QUALITY; RESTORATION;

D O I：

10.1136/bjo-2024-325403

中图分类号：

R77 [眼科学];

学科分类号：

100212 ;

摘要：

Background/aims The aim of this study was to develop and evaluate digital ray, based on preoperative and postoperative image pairs using style transfer generative adversarial networks (GANs), to enhance cataractous fundus images for improved retinopathy detection. Methods For eligible cataract patients, preoperative and postoperative colour fundus photographs (CFP) and ultra-wide field (UWF) images were captured. Then, both the original CycleGAN and a modified CycleGAN (C2ycleGAN) framework were adopted for image generation and quantitatively compared using Frechet Inception Distance (FID) and Kernel Inception Distance (KID). Additionally, CFP and UWF images from another cataract cohort were used to test model performances. Different panels of ophthalmologists evaluated the quality, authenticity and diagnostic efficacy of the generated images. Results A total of 959 CFP and 1009 UWF image pairs were included in model development. FID and KID indicated that images generated by C2ycleGAN presented significantly improved quality. Based on ophthalmologists' average ratings, the percentages of inadequate-quality images decreased from 32% to 18.8% for CFP, and from 18.7% to 14.7% for UWF. Only 24.8% and 13.8% of generated CFP and UWF images could be recognised as synthetic. The accuracy of retinopathy detection significantly increased from 78% to 91% for CFP and from 91% to 93% for UWF. For retinopathy subtype diagnosis, the accuracies also increased from 87%-94% to 91%-100% for CFP and from 87%-95% to 93%-97% for UWF. Conclusion Digital ray could generate realistic postoperative CFP and UWF images with enhanced quality and accuracy for overall detection and subtype diagnosis of retinopathies, especially for CFP.\

引用

页码：1423 / 1429

页数：7

共 50 条

[1] Automated detection of crystalline retinopathy via fundus photography using multistage generative adversarial networks
Choi, Eun Young
Han, Seung Hoon
Ryu, Ik Hee
Kim, Jin Kuk
Lee, In Sik
Han, Eoksoo
Kim, Hyungsu
Choi, Joon Yul
Yoo, Tae Keun
BIOCYBERNETICS AND BIOMEDICAL ENGINEERING, 2023, 43 (04) : 725 - 735
[2] Style Transfer of Urban Road Images Using Generative Adversarial Networks With Structural Details
Li, Yaochen
Wu, Xiao
Lu, Danhui
Li, Ling
Liu, Yuehu
Zhu, Li
IEEE MULTIMEDIA, 2020, 27 (03) : 54 - 65
[3] Generating Fundus Fluorescence Angiography Images from Structure Fundus Images Using Generative Adversarial Networks
Li, Wanyue
Kong, Wen
Chen, Yiwei
Wang, Jing
He, Yi
Shi, Guohua
Deng, Guohua
MEDICAL IMAGING WITH DEEP LEARNING, VOL 121, 2020, 121 : 424 - 439
[4] An Enhancing Framework for Botnet Detection Using Generative Adversarial Networks
Yin, Chuanlong
Zhu, Yuefei
Liu, Shengli
Fei, Jinlong
Zhang, Hetong
2018 INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND BIG DATA (ICAIBD), 2018, : 228 - 234
[5] Semantic Segmentation of Agricultural Images Based on Style Transfer Using Conditional and Unconditional Generative Adversarial Networks
Madokoro, Hirokazu
Takahashi, Kota
Yamamoto, Satoshi
Nix, Stephanie
Chiyonobu, Shun
Saruta, Kazuki
Saito, Takashi K.
Nishimura, Yo
Sato, Kazuhito
APPLIED SCIENCES-BASEL, 2022, 12 (15):
[6] Using Vehicle Synthesis Generative Adversarial Networks to Improve Vehicle Detection in Remote Sensing Images
Zheng, Kun
Wei, Mengfei
Sun, Guangmin
Anas, Bilal
Li, Yu
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION, 2019, 8 (09)
[7] Heightmap Reconstruction of Macula on Color Fundus Images Using Conditional Generative Adversarial Networks
Tahghighi, Peyman
Zoroofi, Reza A.
Safi, Sare
Ramezani, Alireza
Ahmadieh, Hamid
2021 26TH INTERNATIONAL COMPUTER CONFERENCE, COMPUTER SOCIETY OF IRAN (CSICC), 2021,
[8] Automatic Segmentation of Drusen and Exudates on Color Fundus Images using Generative Adversarial Networks
Engelberts, Jonne
Gonzalez-Gonzalo, Cristina
Sanchez, Clara I.
van Grinsven, Mark J.
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE, 2019, 60 (09)
[9] Inter-dataset performance analysis of generative adversarial networks for optic disc segmentation using digital fundus images
Sharma A.
Agrawal M.
Dutta Roy S.
Gupta V.
Research on Biomedical Engineering, 2023, 39 (04) : 863 - 875
[10] Enhancing network intrusion detection performance using generative adversarial networks
Zhao, Xinxing
Fok, Kar Wai
Thing, Vrizlynn L. L.
COMPUTERS & SECURITY, 2024, 145

← 1 2 3 4 5 →