An Automated Grading System for Detection of Vision-Threatening Referable Diabetic Retinopathy on the Basis of Color Fundus Photographs
The DOI matches its Crossref record; 40 of the 40 references it cites were found in a registry or an index.
Abstract
OBJECTIVE: The goal of this study was to describe the development and validation of an artificial intelligence–based, deep learning algorithm (DLA) for the detection of referable diabetic retinopathy (DR). RESEARCH DESIGN AND METHODS: A DLA using a convolutional neural network was developed for automated detection of vision-threatening referable DR (preproliferative DR or worse, diabetic macular edema, or both). The DLA was tested by using a set of 106,244 nonstereoscopic retinal images. A panel of ophthalmologists graded DR severity in retinal photographs included in the development and internal validation data sets (n = 71,043); a reference standard grading was assigned once three graders achieved consistent grading outcomes. For external validation, we tested our DLA using 35,201 images of 14,520 eyes (904 eyes with any DR; 401 eyes with vision-threatening referable DR) from population-based cohorts of Malays, Caucasian Australians, and Indigenous Australians. RESULTS: Among the 71,043 retinal images in the training and validation data sets, 12,329 showed vision-threatening referable DR. In the internal validation data set, the area under the curve (AUC), sensitivity, and specificity of the DLA for vision-threatening referable DR were 0.989, 97.0%, and 91.4%, respectively. Testing against the independent, multiethnic data set achieved an AUC, sensitivity, and specificity of 0.955, 92.5%, and 98.5%, respectively. Among false-positive cases, 85.6% were due to a misclassification of mild or moderate DR. Undetected intraretinal microvascular abnormalities accounted for 77.3% of all false-negative cases. CONCLUSIONS: This artificial intelligence–based DLA can be used with high accuracy in the detection of vision-threatening referable DR in retinal images. This technology offers potential to increase the efficiency and accessibility of DR screening programs.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The classification based on fundus images for referable DR followed the guidelines of the National Health Service (NHS) diabetic eye screening (Supplementary Table 5)40.”
From Economic evaluation for medical artificial intelligence: accuracy vs. cost-effectiveness in a diabetic retinopathy screening case (2024) · cites this paper as [40]
In the citing paper
“We employed a validated AI model for automatic DR screening using fundus images from the Lifeline Express Program and evaluated the AI’s performance against manual grading results40.”
From Economic evaluation for medical artificial intelligence: accuracy vs. cost-effectiveness in a diabetic retinopathy screening case (2024) · cites this paper as [40]
In the citing paper
“It should be noted that the specificity of the AI model in this real-world dataset was lower than the reported performance in its previous validation as well as other reported AI models40.”
From Economic evaluation for medical artificial intelligence: accuracy vs. cost-effectiveness in a diabetic retinopathy screening case (2024) · cites this paper as [40]
In the citing paper
“Li et al. [43] developed a deep CNN-based DR screening system that achieved performance comparable to or exceeding that of professional ophthalmologists across real-world datasets from multiple countries.”
From Artificial intelligence-powered prediction of diabetic complications: from clinical data to molecular omics (2026) · cites this paper as [43]
In the citing paper
“Only eight studies in ophthalmology imaging14,21,32,33,43,55,108,109, ten studies in respiratory imaging64,66,70,72,75,79,82,87,89,110 and six studies in breast imaging62,91,97,104,106,111 mentioned adherence to the STARD-2015 guidelines or had a STARD flow diagram in the manuscript.”
From Diagnostic accuracy of deep learning in medical imaging: a systematic review and meta-analysis (2021) · cites this paper as [33]
Citations
Cites
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Rethinking the Inception Architecture for Computer Vision
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.308 · Cited by 32,138
Dermatologist-level classification of skin cancer with deep neural networks
Esteva, A., Kuprel, B., Novoa, R. A., Ko, J., Swetter, S. M., Blau, H. M., & Thrun, S. (2017) · Nature · doi:10.1038/nature21056 · Cited by 14,362 · cited 1×
“These deep learning techniques have been applied in highly image-driven medical specialties, including dermatology and radiology, with promising results: the techniques have identified diseases as accurately as or more accurately than board-certified specialists (15,16).”
BackgroundInternational Diabetes Federation 2017
Carracher, A. M., Marathe, P. H., & Close, K. L. (2018) · Journal of Diabetes · doi:10.1111/1753-0407.12644 · Cited by 13,418
Gulshan, V., Peng, L., Coram, M., Stumpe, M. C., Wu, D., Narayanaswamy, A., Venugopalan, S., Widner, K., Madams, T., Cuadros, J., et al. (2016) · JAMA · doi:10.1001/jama.2016.17216 · Cited by 7,697
(2007) · Developments in Ophthalmology · doi:10.1159/isbn.978-3-318-01439-6 · Cited by 5,583
Global prevalence and major risk factors of diabetic retinopathy
Yau, J. W. Y., Rogers, S. L., Kawasaki, R., Lamoureux, E. L., Kowalski, J. W., Bek, T., Chen, S.-J., Dekker, J. M., Fletcher, A., Grauslund, J., et al. (2012) · Diabetes Care · doi:10.2337/dc11-1909 · Cited by 4,985 · cited 2×
“6%, corresponding to nearly 100 million people worldwide (1).”
BackgroundSelf-citationZhou, B., Lu, Y., Hajifathalian, K., Bentham, J., Di Cesare, M., Danaei, G., Bixby, H., Cowan, M., Ali, M., Taddei, C., et al. (2016) · The Lancet · doi:10.1016/s0140-6736(16)00618-8 · Cited by 4,311
Wilkinson, C. P., Ferris, F. L., Klein, R. E., Lee, P. P., Agardh, C. D., Davis, M., Dills, D., Kampik, A., Pararajasegaram, R., & Verdaguer, J. T. (2003) · Ophthalmology · doi:10.1016/s0161-6420(03)00475-5 · Cited by 3,743
Self-citationCheung, N., Mitchell, P., & Wong, T. Y. (2010) · The Lancet · doi:10.1016/s0140-6736(09)62124-3 · Cited by 3,288
Self-citation
Cited by
Items 101–200 of 267
Can Artificial Intelligence Make Screening Faster, More Accurate, and More Accessible?
Li, Z., Keel, S. P., & He, M. (2018) · Asia-Pacific Journal of Ophthalmology · doi:10.22608/apo.2018438 · Cited by 29
Self-citationZhao, X., Lin, Z., Yu, S., Xiao, J., Xie, L., Xu, Y., Tsui, C., Cui, K., Zhao, L., Zhang, G., et al. (2022) · Cell Reports Medicine · doi:10.1016/j.xcrm.2023.101197 · Cited by 33
Mamo, A. A., Gebresilassie, B. G., Mukherjee, A., Hassija, V., & Chamola, V. (2024) · Cognitive Computation · doi:10.1007/s12559-024-10291-3 · Cited by 32
Chu, A., Squirrell, D., Phillips, A. M., & Vaghefi, E. (2020) · Journal of Ophthalmology · doi:10.1155/2020/8841927 · Cited by 29
“Ten studies used more than one camera type to take colour fundus photographs for the validation datasets, whilst the known input image resolutions used amongst the studies were 299× 299 pixels [24, 32, 35, 37], 512× 512 pixels [38, 39, 42], and 779× 779 pixels [33, 34].”
BackgroundMethodInfluentialAlam, M. N., Yamashita, R., Ramesh, V., Prabhune, T., Lim, J. I., Chan, R. V. P., Hallak, J., Leng, T., & Rubin, D. L. (2023) · Scientific Reports · doi:10.1038/s41598-023-33365-y · Cited by 30
“The FundusNet model pretrained with style transfer augmentation achieved an average area under the receiver operating characteristics (ROC) curve (AUC) of 0.91 on the independent test dataset from UIC, outperforming the state-of-the-art supervised baseline models (ResNet50 and InceptionV3) trained with Imagenet weight…”
MethodInfluentialDistributed training of foundation models for ophthalmic diagnosis
Gholami, S., Jannat, F.-E., Thompson, A. C., Ong, S. S. Y., Lim, J. I., Leng, T., Tabkhivayghan, H., & Alam, M. N. (2025) · Communications Engineering · doi:10.1038/s44172-025-00341-5 · Cited by 30
Tseng, V. S. M., Chen, C., Liang, C.-M., Tai, M., Liu, J.-T., Wu, P.-Y., Deng, M.-S., Lee, Y.-W., Huang, T., & Chen, Y.-H. (2020) · Translational Vision Science & Technology · doi:10.1167/tvst.9.2.41 · Cited by 28
The Validation of Deep Learning-Based Grading Model for Diabetic Retinopathy
Zhang, W.-F., Li, D., Wei, Q., Ding, D., Meng, L., Wang, Y., Zhao, X., & Chen, Y. (2022) · Frontiers in Medicine · doi:10.3389/fmed.2022.839088 · Cited by 32
Chia, M. A., Hersch, F. B., Sayres, R., Bavishi, P., Tiwari, R., Keane, P. A., & Turner, A. W. (2023) · British Journal of Ophthalmology · doi:10.1136/bjo-2022-322237 · Cited by 27
Machine Learning-Based Diabetic Retinopathy Detection: A Comprehensive Study Using InceptionV3 Model
Deshpande, G., Govardhan, Y., & Jain, A. (2024) · ASU International Conference in Emerging Technologies for Sustainability and Intelligent Systems (ICETSIS) · doi:10.1109/icetsis61505.2024.10459541 · Cited by 25
Prathibha, S., & Siddappaji (2024) · Global Transitions · doi:10.1016/j.glt.2024.04.001 · Cited by 29
Differentiate cavernous hemangioma from schwannoma with artificial intelligence (AI)
Bi, S., Chen, R., Zhang, K., Xiang, Y., Wang, R., Lin, H., & Yang, H. (2020) · Annals of Translational Medicine · doi:10.21037/atm.2020.03.150 · Cited by 25
“In recent years, the application of artificial intelligence (AI) in medicine has achieved physician-equivalent classification accuracy in the diagnosis of many diseases, including diabetic retinopathy (10-13), lung diseases (14), cardiovascular disease (15), liver disease, skin cancer (16), and thyroid cancer (17) and…”
BackgroundTransparency in Artificial Intelligence Reporting in Ophthalmology-A Scoping Review
Chen, D. K., Geevarghese, A., Lee, S., Plovnick, C., Elgin, C. Y., Zhou, R. W., Oermann, E. K., Aphinyonaphongs, Y., & Al‐Aswad, L. A. (2024) · Ophthalmology Science · doi:10.1016/j.xops.2024.100471 · Cited by 27
Yu, M., & Wang, Y. (2022) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.22734 · Cited by 25
“Li et al.(21) Inception v3 Chinese color fundus images Multi-racial color fundus images SE 97.”
BackgroundDiabetic Retinopathy Grading Based on a Hybrid Deep Learning Model
AbdelMaksoud, E., Barakat, S., & Elmogy, M. M. (2020) · International Conference on Data Analytics for Business and Industry (ICDABI) · doi:10.1109/icdabi51230.2020.9325672 · Cited by 22
“EX takes the same color and shape of the Optic Disc (OD) [2].”
BackgroundInfluentialJi, Y., Chen, N., Liu, S., Yan, Z., Qian, H., Zhu, S., Zhang, J., Wang, M., Jiang, Q., & Yang, W. (2022) · Disease Markers · doi:10.1155/2022/3406890 · Cited by 23
“[14] Detection 35201 images Deep learning algorithm The AUC value, sensitivity, and specificity were 0.”
BackgroundMethodArtificial intelligence in retinal imaging: current status and future prospects
Heger, K. A., & Waldstein, S. M. (2023) · Expert Review of Medical Devices · doi:10.1080/17434440.2023.2294364 · Cited by 24
“The AUC for referable DR ranged from 89% to 99% in the individual studies [52–55].”
BackgroundZhang, G., Lin, J., Wang, J., Ji, J., Cen, L., Chen, W., Xie, P., Zheng, Y., Xiong, Y., Wu, H., et al. (2022) · BMJ Open · doi:10.1136/bmjopen-2021-060155 · Cited by 21
Xie, L. Y., Yang, S., Squirrell, D., & Vaghefi, E. (2020) · PLoS ONE · doi:10.1371/journal.pone.0225015 · Cited by 23
“Our extensive investigation (to be published soon as a systematic review) has found only a few published research-based AIs that could be closer to clinical translation [4, 16, 18, 19, 26, 36–38].”
BackgroundDeep-Learning–Based Pre-Diagnosis Assessment Module for Retinal Photographs: A Multicenter Study
Yuen, V. L., Ran, A. R., Shi, J., Sham, K., Yang, D., Chan, V. T. T., Chan, R. N. C., Yam, J. C. S., Tham, C. C. Y., McKay, G. J., et al. (2021) · Translational Vision Science & Technology · doi:10.1167/tvst.10.11.16 · Cited by 20
Self-citationResearch progress and application of artificial intelligence in thyroid associated ophthalmopathy
Diao, J., Chen, X., Shen, Y., Li, J., Chen, Y., He, L., Chen, S., Mou, P., Ma, X., & Wei, R.-L. (2023) · Frontiers in Cell and Developmental Biology · doi:10.3389/fcell.2023.1124775 · Cited by 21
“In ophthalmology, many AI applications have achieved satisfactory performance in diagnosing and predicting several retinal diseases based on the contribution of widely used fundus images (Li et al., 2018b; Nagasato et al., 2018; Peng et al., 2019).”
BackgroundArtificial Intelligence and Deep Learning in Ophthalmology
Wang, Z., Keane, P. A., Chiang, M. F., Cheung, C. Y.-L., Wong, T. Y., & Ting, D. S. W. (2021) · Springer eBooks · doi:10.1007/978-3-030-58080-3_200-1 · Cited by 30
Self-citationGlycemic status and its association with retinal age gap: Insights from the UK biobank study
Chen, R., Xu, J., Zhang, X., Zhang, J., Shang, X., Ge, Z., He, M., Wang, W., & Zhu, Z. (2023) · Diabetes Research and Clinical Practice · doi:10.1016/j.diabres.2023.110817 · Cited by 19
Self-citationVaghefi, E., Yang, S., Xie, L. Y., Hill, S., Schmiedel, O., Murphy, R., & Squirrell, D. (2020) · Diabetic Medicine · doi:10.1111/dme.14386 · Cited by 21
Cao, S., Zhang, R., Jiang, A., Kuerban, M., Wumaier, A., Wu, J., Xie, K., Aizezi, M., Tuersun, A., Liang, X., et al. (2023) · BioMedical Engineering OnLine · doi:10.1186/s12938-023-01097-9 · Cited by 21
“However, for AI-based DR diagnostic systems, screening guidelines suggest that sensitivity and specificity should be more than 80% [18].”
BackgroundZhao, M., & Jiang, Y. (2019) · Eye · doi:10.1038/s41433-019-0629-2 · Cited by 19
Moannaei, M., Jadidian, F., Doustmohammadi, T., Kiapasha, A. M., Bayani, R., Rahmani, M., Jahanbazy, M. R., Sohrabivafa, F., Asadi Anar, M., Magsudy, A., et al. (2025) · BioMedical Engineering OnLine · doi:10.1186/s12938-025-01336-1 · Cited by 18
“…being overestimated in real-world scenarios, and this factor should be considered [19] It is reasonable for another complementary study to initiate using this algorithm in primary care settings; other applications and research are necessary to improve the this algorithm’s clinical validity [24].”
InfluentialDiabetic Retinopathy Diagnosis Through Computer-Aided Fundus Image Analysis: A Review
Kaur, J., Mittal, D., & Singla, R. P. (2021) · Archives of Computational Methods in Engineering · doi:10.1007/s11831-021-09635-1 · Cited by 18
“Deep learning using convolutional neural network for detection of DR and non-DR is developed by Li et al. [128].”
MethodWang, Y., Shi, D., Tan, Z., Niu, Y. Y., Jiang, Y., Xiong, R., Peng, G., & He, M. (2021) · Frontiers in Medicine · doi:10.3389/fmed.2021.740987 · Cited by 23
“Detailed descriptions of this DLA have been reported in our previous study (19).”
BackgroundMethodInfluentialSelf-citationDeep learning-based risk classification and auxiliary diagnosis of macular edema
Wu, T., Liu, L., Zhang, T., & Wu, X. (2022) · Intelligence-Based Medicine · doi:10.1016/j.ibmed.2022.100053 · Cited by 18
Pathological Myopia Image Recognition Strategy Based on Data Augmentation and Model Fusion
Cui, J., Zhang, X., Xiong, F., & Chen, C. (2021) · Journal of Healthcare Engineering · doi:10.1155/2021/5549779 · Cited by 17
“…and diagnostic results are heavily dependent on the doctor’s knowledge and clinical experience, which makes diagnosis subjective and time-consuming [24, 25]. development of medical AI has significantly improved the efficiency of ophthalmic disease diagnosis in clinical work and reduced the burden on…”
BackgroundA deep learning-based ADRPPA algorithm for the prediction of diabetic retinopathy progression
Wang, V. Y., Lo, M., Chen, T., Huang, C.-H., Huang, A., & Wang, P. (2024) · Scientific Reports · doi:10.1038/s41598-024-82884-9 · Cited by 16
Li, L., Ren, P., Lu, Q., Zhou, E., Yu, W., Huang, J., He, X., & Han, W. (2021) · Annals of Translational Medicine · doi:10.21037/atm-20-3275 · Cited by 19
“The false-positive and false-negative images of the validation data set were classified by 5 experienced retinal specialists (28).”
Use of artificial intelligence in ophthalmology: a narrative review
Martins, T. G. D. S., Schor, P., Mendes, L. G. A., Fowler, S. B., & da Silva, R. M. (2022) · Sao Paulo Medical Journal · doi:10.1590/1516-3180.2021.0713.r1.22022022 · Cited by 16
Classification of Color Fundus Photographs Using Fusion Extracted Features and Customized CNN Models
Wang, J., Lu, N., Du, W.-C., Liu, K., Hsu, S., Wang, C.-Y., Chen, Y., Chang, L.-C., Twan, W., Chen, T., et al. (2023) · Healthcare · doi:10.3390/healthcare11152228 · Cited by 16
“It is also essential to assess the cost-effectiveness and the practicality of integrating such AI systems into existing healthcare infrastructures [13–15].”
BackgroundGautam, A., & Shanker, R. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.110715 · Cited by 17
Zhu, A., Tailor, P., Verma, R., Zhang, I., Schott, B., Ye, C., Szirth, B., Habiel, M. M., & Khouri, A. S. (2023) · Journal of Telemedicine and Telecare · doi:10.1177/1357633x231158832 · Cited by 18
“Several studies have retrospectively validated DLAI performance in detecting AMD, 14,24 DR, 32,33 and glaucoma 8 using mydriatic retinal fundus image datasets.”
InfluentialOphthalmology and the emergence of artificial intelligence
Scheetz, J., He, M., & van Wijngaarden, P. (2021) · The Medical Journal of Australia · doi:10.5694/mja2.50932 · Cited by 15
Self-citationInstant automatic diagnosis of diabetic retinopathyPreprint
Quellec, G., Lamard, M., Laÿ, B., Guilcher, A. L., Erginay, A., Cochener, B., & Massin, P. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1906.11875 · Cited by 16
Chen, K.-Y., Chan, H.-C., & Chan, C.-M. (2026) · npj Digital Medicine · doi:10.1038/s41746-026-02627-0 · Cited by 15
“StandardizedreportingofAIoperating thresholdsandreference-standardde fi nitionswillfurtherimproveevidence synthesis and comparability across studies 61 .”
Kim, S. Y., Park, J., Son, Y., Lee, H., Woo, S., Lee, M., Lee, H., Sang, H., Yon, D. K., & Rhee, S. Y. (2025) · JMIR Medical Informatics · doi:10.2196/58107 · Cited by 13
“Li et al [17] also validated a DL system with 35,201 images, obtaining an AUROC of 0.955.”
InfluentialRapid and accurate classification of mung bean seeds based on HPMobileNet
Song, S., Chen, Z., Yu, H., Xue, M., & Liu, J. (2025) · Frontiers in Plant Science · doi:10.3389/fpls.2024.1474906 · Cited by 13
“Through computer vision technology, we can quickly and accurately identify the variety, color, area, size, and other characteristics of mung beans for automated classi fi cation and screening (Li et al., 2018).”
Retinal vascular geometry detection as a biomarker in diabetes mellitus
Li, M., Wang, G., Xia, H., Feng, Z., Xiao, P., & Yuan, J. (2021) · European Journal of Ophthalmology · doi:10.1177/11206721211033488 · Cited by 13
On Detection of Diabetic Retinopathy via Multiple Instance Learning
Avolio, M., Fuduli, A., Vocaturo, E., & Zumpano, E. (2023) · International Database Engineering and Applications Symposium (IDEAS) · doi:10.1145/3589462.3589490 · Cited by 13
“The presence or occurrence of one, two, or all of these characteristics in the retina determines the stage of DR [18].”
BackgroundEyes as the windows into cardiovascular disease in the era of big data
Chan, Y. K., Cheng, C., & Sabanayagam, C. (2023) · Taiwan Journal of Ophthalmology · doi:10.4103/tjo.tjo-d-23-00018 · Cited by 16
“For example, although inception‑v3 was not developed for medical image recognition initially, it has been successfully used for classifying medical images based on transfer learning methods, [25,26] which include the classification of RFP. [4,24,27] Training datasets Few sufficiently large datasets include both eye an…”
MethodTowards population-independent, multi-disease detection in fundus photographs
Matta, S. C., Lamard, M., Conze, P.-H., Guilcher, A. L., Lecat, C., Carette, R., Basset, F., Massin, P., Rottier, J.-B., Cochener, B., et al. (2023) · Scientific Reports · doi:10.1038/s41598-023-38610-y · Cited by 14
Wang, Y., Han, X., Li, C., Luo, L., Yin, Q., Zhang, J., Peng, G., Shi, D., & He, M. (2024) · Journal of Medical Internet Research · doi:10.2196/52506 · Cited by 19
“The original version of the algorithm has been reported previously [19].”
Self-citationImaging the eye and its relevance to diabetes care
Quinn, N. B., Jenkins, A. J., Ryan, C. D., Januszewski, A. S., Pető, T., & Brazionis, L. (2020) · Journal of Diabetes Investigation · doi:10.1111/jdi.13462 · Cited by 13
“Li et al.(29) developed and validated an AI system for the detection of sightthreatening referable DR, producing a sensitivity and specificity of 92.”
BackgroundRahmati, M., Smith, L., Piyasena, M. P., Bowen, M., Boyer, L., Fond, G., Kazemi, A., Yon, D. K., Lee, H., Sehmbi, T., et al. (2025) · Eye · doi:10.1038/s41433-025-03849-4 · Cited by 11
“Assessment of DR screening images by AI, instead of by scarce trained specialists could potentially increases the efficiency and accessibility of screening programs, and may even decrease cost of screening [14].”
Chen, Y., Song, F., Zhao, Z., Wang, Y., To, E., Liu, Y., Shi, D., Chen, X., Xu, L., Shang, X., et al. (2025) · Diabetes Research and Clinical Practice · doi:10.1016/j.diabres.2025.112161 · Cited by 13
Self-citationAlshammari, R. D., Haliqa, W. A., & Majzoub, R. Y. (2025) · Journal of Medical Imaging and Interventional Radiology · doi:10.1007/s44326-025-00083-w · Cited by 11
“• DIARETDB1 (around 89 images): Focused on the DR lesions of early stages [40].”
Brazionis, L., Quinn, N. B., Dabbah, S., Ryan, C. D., Møller, D., Richardson, H., KEECH, A. C., Januszewski, A. S., Grauslund, J., Rasmussen, M. L., et al. (2023) · Graefe s Archive for Clinical and Experimental Ophthalmology · doi:10.1007/s00417-023-06002-7 · Cited by 11
Nishimura, K., Tabuchi, H., Nakakura, S., Nakatani, Y., Yorihiro, A., Hasegawa, S., Tanabe, H., Noguchi, A., Aoki, R., & Kiuchi, Y. (2019) · Translational Vision Science & Technology · doi:10.1167/tvst.8.3.55 · Cited by 11
Cano, J., O’Neill, W. D., Penn, R. D., Blair, N. P., Kashani, A. H., Ameri, H., Kaloostian, C. L., & Shahidi, M. (2020) · Biomedical Optics Express · doi:10.1364/boe.394472 · Cited by 12
Artificial Intelligence in Diabetic Retinopathy Screening. A Review
Straňák, Z., Penčák, M., & Veith, M. (2021) · Czech and Slovak Ophthalmology · doi:10.31348/2021/6 · Cited by 12
Diabetes management in the era of artificial intelligence
Papazafiropoulou, A. K. (2024) · Archives of Medical Science - Atherosclerotic Diseases · doi:10.5114/amsad/183420 · Cited by 16
“An unnamed AI algorithm was created by Li et al. [44].”
MethodA Deep Learning Segmentation Model for Detection of Active Proliferative Diabetic Retinopathy
Dinesen, S., Schou, M. G., Hedegaard, C. V., Subhi, Y., Savarimuthu, T. R., Pető, T., Andersen, J., & Grauslund, J. (2025) · Ophthalmology and Therapy · doi:10.1007/s40123-025-01127-w · Cited by 10
“They also investigated the classification of vision-threatening DR with DL defined as severe non-proliferative DR (ICDR level 3) or above, achieving sensitivities from 95.2% to 100% and specificities from 89.5% to 97.4% [11, 12, 15, 16, 27].”
InfluentialBildgebung der diabetischen Retinopathie
Ziemssen, F., Roeck, D., Marahrens, L., & Agostini, H. T. (2018) · Der Diabetologe · doi:10.1007/s11428-018-0410-3 · Cited by 10
Deep learning detection of diabetic retinopathy in Scotland’s diabetic eye screening programme
Fleming, A., Mellor, J., McGurnaghan, S. J., Blackbourn, L. A. K., Goatman, K. A., Styles, C., Storkey, A., McKeigue, P., & Colhoun, H. M. (2023) · British Journal of Ophthalmology · doi:10.1136/bjo-2023-323395 · Cited by 10
Rautaray, J., Ali, A. B. M., Kandpal, M., Mishra, P., Rashid, R. F., Alimova, F., Kallel, M., & Batool, N. (2025) · SLAS TECHNOLOGY · doi:10.1016/j.slast.2025.100325 · Cited by 10
RetractedWang, N., Liu, X., Fang, X., Chen, S., Xi, Z., Zhang, X., Xue, C., Liu, X., & Xie, L. (2022) · Disease Markers · doi:10.1155/2022/9515571 · Cited by 10
“Machine learning is a powerful tool to perform complex algorithms to detect and diagnose clinical diseases [42, 43].”
BackgroundRetractedDeep learning: applications in retinal and optic nerve diseases
Charng, J., Alam, K., Swartz, G., Kugelman, J., Alonso‐Caneiro, D., Mackey, D. A., & Chen, F. K. (2022) · Clinical and Experimental Optometry · doi:10.1080/08164622.2022.2111201 · Cited by 13
“Two out of the five algorithms reported sensitivity in detecting DR in the high 90s (Table 1) 30,32–36 with higher specificity reported by the manufacturers of RetinoScan 34 than ML based Retmarker 35 (92 vs 63).”
Trends and hotspots in the field of diabetic retinopathy imaging research from 2000–2023
Zhang, Q., Zhang, P., Chen, N.-M., Zhu, Z., Li, W., & Wang, Q. (2024) · Frontiers in Medicine · doi:10.3389/fmed.2024.1481088 · Cited by 9
He, S., Bulloch, G., Zhang, L., Xie, Y., Wu, W., He, Y., Meng, W., Shi, D., & He, M. (2023) · Current Eye Research · doi:10.1080/02713683.2023.2215984 · Cited by 10
“31 Although our model achieved good performance upon external validation this was tested by a dataset using the same camera for image capture 6,7,18 and until now these algorithms were only assumed to generalize well across camera models.”
BackgroundMethodSelf-citationGogulamudi, N., Madhu, G., Naidu, G. A., Kautish, S., Almazyad, A. S., Xiong, G., & Mohamed, A. W. (2024) · Results in Engineering · doi:10.1016/j.rineng.2024.102820 · Cited by 9
The Role of Artificial Intelligence in the Diagnosis and Management of Diabetic Retinopathy
Ansari, A., Ansari, N. M., Khalid, U., Markov, D., Bechev, K. I., Aleksiev, V., Markov, G., & Poryazova, E. G. (2025) · Journal of Clinical Medicine · doi:10.3390/jcm14145150 · Cited by 12
“Systems like IDx-DR and EyeArt leverage large annotated datasets and hierarchical feature learning to automatically identify DR with high precision [34–36].”
Li, Y., Sheu, W. H., Chou, C., Lin, C.-H., Cheng, Y.-S., Wang, C.-Y., Wu, C., & Lee, I. (2021) · Life · doi:10.3390/life11030200 · Cited by 9
“Recent publications have shown that the diagnostic performance of some deep learning algorithms is similar to, or even better than, that of human experts [12,13].”
ResultQurban, Q., & Cassidy, L. (2024) · SAGE Open Medicine · doi:10.1177/20503121241274197 · Cited by 8
Zhao, X., Gu, X., Teng, D., Sun, X., Wei, Q., Wang, B., Wang, J., Zhao, J., Ding, D., Zhang, B., et al. (2025) · Cell Reports Medicine · doi:10.1016/j.xcrm.2025.102187 · Cited by 8
“Firstly, most of the existing studies used conventional fundus images to develop DLAs, 17,18 but these systems could not detect peripheral retinal and vascular abnormalities, limiting their utility for comprehensive fundus screening.”
Jan, C., Joseph, S., Vingrys, A. J., Henwood, J., Ge, Z., Stafford, R. S., & He, M. (2025) · npj Digital Medicine · doi:10.1038/s41746-025-01768-y · Cited by 8
“The development and validation procedures for each algorithm have been documented in prior literature, with detailed references provided herein 16,25,26 .”
Self-citationArtificial Intelligence-Based Screening System for Diabetic Retinopathy in Primary Care
Baget‐Bernaldiz, M., Póveda, B. F., Romero‐Aroca, P., Navarro-Gil, R., Hernando-Comerma, A., Bautista-Pérez, Á., Llagostera-Serra, M., Morente-Lorenzo, C., Vizcarro, M., & Mira-Puerto, A. (2024) · Diagnostics · doi:10.3390/diagnostics14171992 · Cited by 10
“The S and SP were found to be 92.5% and 98.5%, respectively [21].”
BackgroundInfluentialJoseph, S., Wang, Y., Drinkwater, J. J., Jan, C. L., Sundar, B., Zhu, Z., Shang, X., Henwood, J., Kiburg, K., Clark, M., et al. (2025) · British Journal of Ophthalmology · doi:10.1136/bjo-2025-327447 · Cited by 9
Self-citationCapacity building in screening and treatment of diabetic retinopathy in Asia-Pacific region
Silpa‐archa, S., Limwattanayingyong, J., Tadarati, M., Amphornphruet, A., & Ruamviboonsuk, P. (2021) · Indian Journal of Ophthalmology · doi:10.4103/ijo.ijo_1075_21 · Cited by 7
Diabetic Retinopathy Images Classification via Multiple Instance Learning
Vocaturo, E., & Zumpano, E. (2021) · IEEE/ACM Conference on Connected Health: Applications, Systems and Engineering Technologies · doi:10.1109/chase52844.2021.00034 · Cited by 7
Han, X., Jiang, Y., Niu, Y. Y., Zhu, Y., Huang, W., & He, M. (2021) · British Journal of Ophthalmology · doi:10.1136/bjophthalmol-2020-318252 · Cited by 7
Self-citationŠtotl, I., Blagus, R., & Urbančič‐Rovan, V. (2021) · Diabetologia · doi:10.1007/s00125-021-05604-2 · Cited by 7
“Clinical prediction models have been used in screening for diabetes in the general population [19] and for automated staging of diabetic retinopathy [20].”
MethodArtificial intelligence in ophthalmology: from innovation to clinical integration
Gurnani, B., & Kaur, K. (2026) · Frontiers in Ophthalmology · doi:10.3389/fopht.2026.1839194 · Cited by 7
Ocular biomarkers: useful incidental findings by deep learning algorithms in fundus photographs
Martin, E., Cook, A. G., Frost, S. M., Turner, A. W., Chen, F. K., McAllister, I. L., Nolde, J. M., & Schlaich, M. P. (2024) · Eye · doi:10.1038/s41433-024-03085-2 · Cited by 8
“This ambiguity appears to increase for earlier shared biomarkers, such as capillary rarefaction and reduced vessel density as seen with OCTA [76, 77] and is demonstrated algorithmically by a 6% specificity rise for Algorithm 2 when mild diabetic retinopathy detection was excluded [68, 69].”
BackgroundMethodInfluentialDash, S. K., Satyanarayana, K., Behera, S. K., Jena, S., Ratha, A. K., Sethy, P. K., & Nanthaamornphong, A. (2025) · Applied Computational Intelligence and Soft Computing · doi:10.1155/acis/6653543 · Cited by 6
“Deep learning (DL) has become a leading technology in the feld of computer vision, garnering substantial research attention for the development of sophisticated medical image processing algorithms aimed at disease detection and diagnosis [6–12].”
AI in Healthcare: Implications for Family Medicine and Primary Care
Wojda, T. R., Hoffman, C. A., Jackson, J., Conti, T., & Maier, J. (2023) · Artificial intelligence · doi:10.5772/intechopen.111498 · Cited by 7
Artificial intelligence: how it works and criteria for assessment
Леонидовна, Ш. И., Garanina, O. E., Klemenova, I. A., Uskova, K. A., Mironycheva, A. M., Dardyk, V. I., & Laskov, V. (2021) · Consilium Medicum · doi:10.26442/20751753.2021.8.201148 · Cited by 7
Performance of Deep Transfer Learning for Detecting Abnormal Fundus Images
Yu, Y., Chen, X., Zhu, X., Zhang, P., Hou, Y., Zhang, R., & Wu, C. (2020) · Journal of Current Ophthalmology · doi:10.4103/joco.joco_123_20 · Cited by 7
AI for the Detection of the Diabetic Retinopathy
Vocaturo, E., & Zumpano, E. (2022) · Internet of things · doi:10.1007/978-3-030-91181-2_8 · Cited by 6
Comprehensive Factors for Predicting the Complications of Diabetes Mellitus: A Systematic Review
Erandathi, M., Wang, W. Y. C., Mayo, M., & Lee, C. (2024) · Current Diabetes Reviews · doi:10.2174/0115733998271863231116062601 · Cited by 8
Abbas, Q., Ibrahim, M. E. A., & Baig, A. R. (2022) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2022.023670 · Cited by 5
“Deep learning (DL) models especially deep convolution neural networks (CNN) have demonstrated outperform performance in the grading of DR severity levels in several settings and on several datasets when compared with traditional hand-designed methods [6].”
BackgroundInfluentialIdentifying the Level of Diabetic Retinopathy Using Deep Convolution Neural Network
Hassan, R., Rahman, M. A., Ullah, I., Alenezi, A. H., & Rassem, T. H. (2020) · Emerging Technology in Computing, Communication and Electronics (ETCCE) · doi:10.1109/etcce51779.2020.9350905 · Cited by 5
“DR classification was performed by ophthalmologists using the DR (ICDR) International Medical Classification methodology [13].”
MethodXie, X., Wu, C., Huang, Z., Zhou, Y., Huang, J., Dao, F., Yan, D., Deng, K., Lyu, H., Ma, C., et al. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag083 · Cited by 9
“Models aligned with screening guidelines report AUC > 0.95 for referable DR in diverse populations [43].”
InfluentialWen, Y., Zeng, Y., Liu, S., Xiong, L., Li, H., Wang, Y., Jia, W., Wang, C., Ma, P., & Liang, Z. C. (2025) · The Visual Computer · doi:10.1007/s00371-025-04187-y · Cited by 5
“Withinthedeeplearn-ing framework, DR grading can generally be regarded as an image classification task, in which features from lesion areas are extracted and analyzed to achieve accurate discrimination of different disease stages [5, 6].”
A Real-World Comparison of Three Deep Learning Systems for Diabetic Retinopathy in Remote Australia
Drinkwater, J. J., Li, Q., Woods, K., Douglas, E., Chia, M. A., Zhou, Y., Bartnik, S., Shah, Y., Shah, V., Keane, P. A., et al. (2025) · Diabetology · doi:10.3390/diabetology6120146 · Cited by 4
“However, Optain had lower performance when only assessing First Nations people [15,16].”
InfluentialUsing deep leaning models to detect ophthalmic diseases: A comparative study
Li, Z., Guo, X., Zhang, J., Liu, X., Chang, R. T., & He, M. (2023) · Frontiers in Medicine · doi:10.3389/fmed.2023.1115032 · Cited by 4
“This process has been described in detail elsewhere (22, 23).”
MethodSelf-citationXu, X., Zhang, J., Song, X., Liu, X., Liu, Y., Feng, L., Su, Y. A., Li, Y., Lu, L., & Fan, X. (2025) · British Journal of Ophthalmology · doi:10.1136/bjo-2025-327442 · Cited by 3
Multimodal Deep Learning for Diabetic Retinopathy: A Survey
Qin, G.-Y., Peng, Q., Zhou, Y.-S., Peng, J., Liu, Q., & Xu, J. (2025) · Journal of Ophthalmology · doi:10.1109/mipr67560.2025.00032 · Cited by 4
“Li et al. [30] proposed an automated grading algorithm on EyePACS fundus photographs that detected vision-threatening DR with 91 % accuracy on independent test sets.”
Enhancing AI-based diabetic retinopathy diagnosis through universal cross-camera image adaptation
Joseph, S., Chen, X., Liu, C., Zhu, Z., Ramasamy, K., Ravilla, T., Ge, Z., & He, M. (2025) · BMJ Open Ophthalmology · doi:10.1136/bmjophth-2025-002238 · Cited by 3
“26 The algorithm was trained on over 200 000 fundus photographs from the EyePACS data set, collected from diverse ophthalmic clinical settings.”
Self-citationConvolutional neural network algorithms in diabetic retinopathy: how far does it go?
Hu, Z., Ji, J., Lin, J.-W., Xiao, C., & Cen, L. (2025) · Artificial Intelligence Review · doi:10.1007/s10462-025-11453-x · Cited by 3
Integrated deep learning model for multi-label retinal disease diagnosis
Mahmood, M. A., Alsalem, K. O., Elbashir, M. K., El-Ghany, S. A., & El-Aziz, A. A. A. (2026) · Scientific Reports · doi:10.1038/s41598-026-54982-3 · Cited by 3
A Label Uncertainty-Guided Multi-Stream Model for Disease Screening
Liu, C., Ge, Z., He, M., & Han, X. (2022) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi52829.2022.9761483 · Cited by 3
“, doctors or trained graders) in a collaborative pattern, where the consistent decision from the majority is often regarded as the “ground truth” of yi [1, 2, 3, 4].”
BackgroundResultSelf-citationMenia, N. K., Diwan, S. G., Mehndiratta, A., & Venkatesh, P. (2024) · Expert Review of Ophthalmology · doi:10.1080/17469899.2024.2328620 · Cited by 3
Diabetic Disease of the Eye in Canada: Consensus Statements from a Retina Specialist Working Group
Omar, A., Williams, G., Whelan, J., Noble, J., Brent, M. H., Giunta, M., Olivier, S., & Lhor, M. (2024) · Ophthalmology and Therapy · doi:10.1007/s40123-024-00923-0 · Cited by 3
“In multiple studies, deep-learning systems for automated DR classification were found to be equal to or better than board-certified specialists [137–141].”
MethodWang, J., Qin, X., Chen, M., Zhang, H.-M., Liu, X., Xiang, M.-J., Hu, J., Huang, H.-Y., & Lan, C. (2024) · International Journal of Ophthalmology · doi:10.18240/ijo.2024.04.02 · Cited by 3
Yu, T. T., Ma, D., Lo, J., Ju, M. J., Beg, M. F., & Šarunic, M. V. (2021) · Biomedical Optics Express · doi:10.1364/boe.431992 · Cited by 3
Metadata
- Authors
- Li, Z., Keel, S., Liu, C., He, Y., Meng, W., Scheetz, J., Lee, P. Y., Shaw, J., Ting, D., Wong, T. Y., Taylor, H., Chang, R., & He, M.
- Affiliations
- Sun Yat-sen University (CN)
- The University of Melbourne (AU)
- Guangzhou Healgoo Interactive Medical Technology Co. Ltd., Guangzhou, China
- Baker Heart and Diabetes Institute (AU)
- National University of Singapore (SG)
- Stanford University (US)
- Corresponding author
- He, Mingguang
- ORCID
- Venue
- Diabetes Care
- Publisher
- American Diabetes Association
- Published
- 2018-10-01
- Volume
- 41
- Issue
- 12
- Pages
- 2509-2516
- ISSN
- 0149-5992 (ISSN-L)
- 1935-5548
- OpenAlex
- W2893365278
- Cited by
- 255
- References
- 39
- Citations by year
- 2026: 23 · 2025: 31 · 2024: 37 · 2023: 43 · 2022: 29 · 2021: 38 · 2020: 31 · 2019: 19 · 2018: 4
- FWCI
- 15.89
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Diabetes Care (publishedVersion)
- PubMed (publishedVersion)
- Funders
- National Natural Science Foundation of China (81420108008)
- Science and Technology Planning Project (2013B20400003)
- Bupa Health Foundation
- Topics
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Retinal Imaging and Analysis
- Health Sciences › Medicine › Ophthalmology › Retinal Diseases and Treatments
- Health Sciences › Medicine › Ophthalmology › Ophthalmology and Visual Health Research
- Keywords
- referable diabetic retinopathy, automated grading, color fundus images, deep learning, convolutional neural networks, diabetic macular edema, preproliferative diabetic retinopathy, retinal image analysis, diabetic retinopathy screening, intraretinal microvascular abnormalities
- MeSH terms
- Adult, Aged, Aged, 80 and over, Algorithms, Artificial Intelligence, Australia, Automation, Diabetic Retinopathy, Diagnosis, Computer-Assisted, Diagnostic Techniques, Ophthalmological, Female, Humans, Male, Mass Screening, Middle Aged, Photography, Sensitivity and Specificity
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.