AI-DRIVEN OCULOMICS: RETINAL IMAGING BIOMARKERS FOR SYSTEMIC DISEASE - FROM CARDIOVASCULAR RISK TO NEURODEGENERATION AND FOUNDATION MODELS
DOI:
https://doi.org/10.31435/ijitss.3(51).2026.5896Keywords:
Oculomics, Retinal Biomarkers, Artificial Intelligence, Deep Learning, Cardiovascular Disease, Foundation ModelsAbstract
Introduction and aim: The retina offers a unique, non-invasive window into systemic vascular and neurological health. Deep-learning algorithms applied to retinal photographs and optical coherence tomography (OCT) can predict cardiovascular events, detect chronic kidney disease, estimate biological age, and identify neurodegenerative diseases - a field termed oculomics. This review synthesizes current evidence on AI-driven oculomics across cardiovascular, cardio-renal-metabolic, neurodegenerative, and aging domains.
Materials and methods: A narrative review was conducted in PubMed, Scopus, and Google Scholar for studies published 2018-2026. Original research, validation studies, and major reviews on AI-based prediction of systemic disease from retinal imaging were included.
Results: Deep-learning models trained on fundus photographs predict major adverse cardiovascular events and ten-year atherosclerotic cardiovascular disease risk with areas under the receiver operating characteristic curve (AUROCs) of 0.70-0.89. Chronic kidney disease detection achieves AUROCs of 0.85-0.94. Alzheimer’s disease classifiers reach AUROCs of 0.93 in development cohorts and 0.73-0.91 in external validation, while Parkinson’s disease detection achieves AUROCs of 0.77-0.92. Retinal age-gap independently predicts mortality. Self-supervised foundation models (RETFound, VisionFM, EyeFound, RETFound-Green, MIRAGE, V-JEPA) show state-of-the-art performance, with volumetric OCT (V-JEPA) reaching AUROCs of 0.94.
Conclusions: AI-driven oculomics is a rapidly maturing field with strong translational potential. Despite encouraging accuracies, prospective validation, algorithmic bias, regulatory approval, and clinical integration remain key challenges. Future work should prioritize multi-ethnic cohort studies, cost-effectiveness analyses, and regulatory pathways.
References
Wagner, S. K., Fu, D. J., Faes, L., Liu, X., Huemer, J., Khalid, H., Ferraz, D., Korot, E., Kelly, C., Balaskas, K., Denniston, A. K., & Keane, P. A. (2020). Insights into systemic disease through retinal imaging-based oculomics. Translational Vision Science & Technology, 9(2), 6. https://doi.org/10.1167/tvst.9.2.6
Zhu, Z., Wang, Y., Qi, Z., Hu, W., Zhang, X., Wagner, S. K., Wang, Y., Ran, A. R., Ong, J., Waisberg, E., Masalkhi, M., Suh, A., Tham, Y. C., Cheung, C. Y., Yang, X., Yu, H., Ge, Z., Wang, W., Sheng, B., . . . Wong, T. Y. (2025). Oculomics: Current concepts and evidence. Progress in Retinal and Eye Research, 106, 101350. https://doi.org/10.1016/j.preteyeres.2025.101350
Zhu, Z., Hu, W., Chen, R., Xiong, R., Wang, W., Shang, X., Chen, Y., Kiburg, K., Shi, D., He, S., Huang, Y., Zhang, X., Tang, S., Zeng, J., Yu, H., Yang, X., & He, M. (2022). Retinal age gap as a predictive biomarker of stroke risk. BMC Medicine, 20(1), 466. https://doi.org/10.1186/s12916-022-02620-w
Ghenciu, L. A., Dima, M., Stoicescu, E. R., Iacob, R., Boru, C., & Haţegan, O. A. (2024). Retinal imaging-based oculomics: Artificial intelligence as a tool in the diagnosis of cardiovascular and metabolic diseases. Biomedicines, 12(9), 2150. https://doi.org/10.3390/biomedicines12092150
Poplin, R., Varadarajan, A. V., Blumer, K., Liu, Y., McConnell, M. V., Corrado, G. S., Peng, L., & Webster, D. R. (2018). Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nature Biomedical Engineering, 2(3), 158–164. https://doi.org/10.1038/s41551-018-0195-0
Rim, T. H., Lee, C. J., Tham, Y. C., Cheung, N., Yu, M., Lee, G., Kim, Y., Ting, D. S. W., Chong, C. C. Y., Choi, Y. S., Yoo, T. K., Ryu, I. H., Baik, S. J., Kim, Y. A., Kim, S. K., Lee, S. H., Lee, B. K., Kang, S. M., Wong, E. Y. M., . . . Wong, T. Y. (2021). Deep-learning-based cardiovascular risk stratification using coronary artery calcium scores predicted from retinal photographs: A multicohort study. The Lancet Digital Health, 3(5), e306-e316. https://doi.org/10.1016/S2589-7500(21)00043-1
Tseng, R. M. W. W., Rim, T. H., Shantsila, E., Yi, J. K., Park, S., Kim, S. S., Lee, C. J., Thakur, S., Nusinovici, S., Peng, Q., Li, H., Xue, C. C., Zhang, L., Cheung, C. Y., Lip, G. Y. H., Wong, T. Y., & Cheng, C. Y. (2023). Validation of a deep-learning-based retinal biomarker (Reti-CVD) in the prediction of cardiovascular disease: Data from UK Biobank. BMC Medicine, 21, 28. https://doi.org/10.1186/s12916-023-02732-x
Diaz-Pinto, A., Ravikumar, N., Attar, R., Suinesiaputra, A., Zhao, Y., Levelt, E., Dall’Armellina, E., Lorenzi, M., Chen, Q., Keenan, T. D. L., Agrón, E., Chew, E. Y., Lu, Z., Gale, C. P., Gale, R. P., Plein, S., & Frangi, A. F. (2022). Predicting myocardial infarction through retinal scans and minimal personal information. Nature Machine Intelligence, 4(1), 55–61. https://doi.org/10.1038/s42256-021-00427-7
Vaghefi, E., Squirrell, D., Yang, S., An, S., Xie, L., Durbin, M. K., Hou, H., Marshall, J., Shreibati, J., McConnell, M. V., & Budoff, M. (2024). Development and validation of a deep-learning model to predict 10-year atherosclerotic cardiovascular disease risk from retinal images using the UK Biobank and EyePACS 10K datasets. American Journal of Preventive Cardiology, 17, 100633. https://doi.org/10.1016/j.ajpc.2023.100633
Chang, J., Ko, A., Park, S. M., Choi, S., Kim, K., Kim, S. M., Yun, J. M., Kang, U., Shin, I. H., Shin, J. Y., Ko, T., Lee, J., Oh, B. L., & Park, K. H. (2020). Association of cardiovascular mortality and deep learning-funduscopic atherosclerosis score derived from retinal fundus images. American Journal of Ophthalmology, 217, 121–130. https://doi.org/10.1016/j.ajo.2020.03.027
Cheung, C. Y., Xu, D., Cheng, C. Y., Sabanayagam, C., Tham, Y. C., Yu, M., Rim, T. H., Chai, C. Y., Gopinath, B., Mitchell, P., Poulton, R., Moffitt, T. E., Caspi, A., Yam, J. C., Tham, C. C., Jonas, J. B., Wang, Y. X., Song, S. J., Burrell, L. M., . . . Wong, T. Y. (2021). A deep-learning system for the assessment of cardiovascular disease risk via the measurement of retinal-vessel calibre. Nature Biomedical Engineering, 5(6), 498–508. https://doi.org/10.1038/s41551-020-00626-4
Zhang, K., Liu, X., Xu, J., Yuan, J., Cai, W., Chen, T., Wang, K., Gao, Y., Nie, S., Xu, X., Qin, X., Su, Y., Xu, W., Olvera, A., Xue, K., Li, Z., Zhang, M., Zeng, X., Zhang, C. L., . . . Wang, G. (2021). Deep-learning models for the detection and incidence prediction of chronic kidney disease and type 2 diabetes from retinal fundus images. Nature Biomedical Engineering, 5(6), 533–545. https://doi.org/10.1038/s41551-021-00745-6
Sabanayagam, C., Xu, D., Ting, D. S. W., Nusinovici, S., Banu, R., Hamzah, H., Lim, C., Tham, Y. C., Cheung, C. Y., Tai, E. S., Wang, Y. X., Jonas, J. B., Cheng, C. Y., Lee, M. L., Hsu, W., & Wong, T. Y. (2020). A deep learning algorithm to detect chronic kidney disease from retinal photographs in community-based populations. The Lancet Digital Health, 2(6), e295-e302. https://doi.org/10.1016/S2589-7500(20)30063-7
Joo, Y. S., Rim, T. H., Koh, H. B., Yi, J., Kim, H., Lee, G., Kim, Y. A., Kang, S. W., Kim, S. S., & Park, J. T. (2023). Non-invasive chronic kidney disease risk stratification tool derived from retina-based deep learning and clinical factors. npj Digital Medicine, 6, 114. https://doi.org/10.1038/s41746-023-00860-5
Wu, Q., Li, J., Zhao, L., Liu, D., Wen, J., Wang, Y., Wang, Y., Huang, N., Jiang, L., Liu, Q., Lin, H., Wan, P., Yang, S., Chen, W., Ye, H., Hassan, M. H. R., Nur, A. H., Dai, Z., Guo, J., . . . Lin, H. (2025). A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images. Nature Communications, 16, 62273. https://doi.org/10.1038/s41467-025-62273-0
Bhak, Y., Lee, Y. H., Kim, J., Lee, K., Lee, D., Jang, E. C., Jang, E., Lee, C. S., Kang, E. S., Park, S., Han, H. W., & Nam, S. M. (2025). Diagnosis of chronic kidney disease using retinal imaging and urine dipstick data: Multimodal deep learning approach. JMIR Medical Informatics, 13, e55825. https://doi.org/10.2196/55825
Yip, M. Y. T., Lim, G., Lim, Z. W., Nguyen, Q. D., Chong, C. C. Y., Yu, M., Bellemo, V., Xie, Y., Lee, X. Q., Hamzah, H., Ho, J., Tan, T. E., Sabanayagam, C., Grzybowski, A., Tan, G. S. W., Hsu, W., Lee, M. L., Wong, T. Y., & Ting, D. S. W. (2020). Technical and imaging factors influencing performance of deep learning systems for diabetic retinopathy. npj Digital Medicine, 3, 40. https://doi.org/10.1038/s41746-020-0247-1
Zhang, L., Yuan, M., An, Z., Zhao, X., Wu, H., Li, H., Wang, Y., Sun, B., Li, H., Ding, S., Zeng, X., Chao, L., Li, P., & Wu, W. (2020). Prediction of hypertension, hyperglycemia and dyslipidemia from retinal fundus photographs via deep learning: A cross-sectional study of chronic diseases in central China. PLOS ONE, 15(5), e0233166. https://doi.org/10.1371/journal.pone.0233166
Cheung, C. Y., Ran, A. R., Wang, S., Chan, V. T. T., Sham, K., Hilal, S., Venketasubramanian, N., Cheng, C. Y., Sabanayagam, C., Tham, Y. C., Schmetterer, L., McKay, G. J., Williams, M. A., Wong, A., Au, L. W. C., Lu, Z., Yam, J. C., Tham, C. C., Chan, J. C., . . . Wong, T. Y. (2022). A deep-learning model for detection of Alzheimer’s disease based on retinal photographs: A retrospective, multicentre case-control study. The Lancet Digital Health, 4(11), e806-e815. https://doi.org/10.1016/S2589-7500(22)00169-8
Wagner, S. K., Hughes, F., Cortina-Borja, M., Pontikos, N., Struyven, R., Liu, X., Montgomery, H., Alexander, D. C., Topol, E., Petersen, S. E., Balaskas, K., Hindley, J., Petzold, A., Rahi, J. S., Denniston, A. K., & Keane, P. A. (2022). AlzEye: Longitudinal record-level linkage of ophthalmic imaging and hospital admissions of 353,157 patients in London, UK. BMJ Open, 12(3), e058552. https://doi.org/10.1136/bmjopen-2021-058552
Tian, J., Smith, G., Guo, H., Liu, B., Pan, Z., Wang, Z., Xiong, S., & Fang, R. (2021). Modular machine learning for Alzheimer’s disease classification from retinal vasculature. Scientific Reports, 11, 238. https://doi.org/10.1038/s41598-020-80312-2
Hao, J., Kwapong, W. R., Shen, T., Fu, H., Xu, Y., Lu, Q., Liu, S., Zhang, J., Liu, Y., Zhao, Y., Zheng, Y., Frangi, A. F., Zhang, S., Qi, H., & Zhao, Y. (2024). Early detection of dementia through retinal imaging and trustworthy AI (Eye-AD). npj Digital Medicine, 7, 294. https://doi.org/10.1038/s41746-024-01304-4
Tham, Y. C., Anees, A., Zhang, L., Goh, J. H. L., Rim, T. H., Nusinovici, S., Hamzah, H., Chee, M. L., Tjio, G., Li, S., Xu, X., Goh, R., Lamoureux, E., Sabanayagam, C., Cheung, C. Y., Tang, F., Wang, Y. X., Nangia, V., Jonas, J. B., . . . Liu, Y. (2021). Referral for disease-related visual impairment using retinal photograph-based deep learning: A proof-of-concept, model development study. The Lancet Digital Health, 3(1), e29-e40. https://doi.org/10.1016/S2589-7500(20)30226-0
Richardson, A., Kundu, A., Henao, R., Lee, T., Scott, B. L., Grewal, D. S., & Fekrat, S. (2024). Multimodal retinal imaging classification for Parkinson’s disease using a convolutional neural network. Translational Vision Science & Technology, 13(8), 23. https://doi.org/10.1167/tvst.13.8.23
Tran, C., Shen, K., Liu, K., Ashok, A., Ramirez-Zamora, A., Chen, J., Li, Y., & Fang, R. (2024). Deep learning predicts prevalent and incident Parkinson’s disease from UK Biobank fundus imaging. Scientific Reports, 14, 3637. https://doi.org/10.1038/s41598-024-54251-1
Sim, M. A., Tham, Y. C., Nusinovici, S., Quek, T. C., Yu, M., Xue, C. C., Chee, M. L., Peng, Q. S., Tan, E. S. J., Chan, S. P., Cai, Y., Chong, E. J. Y., Tan, B. Y., Venketasubramanian, N., Hilal, S., Lai, M. K. P., Choi, H., Richards, A. M., Cheng, C. Y., & Chen, C. L. H. (2025). A deep-learning retinal aging biomarker for cognitive decline and incident dementia. Alzheimer’s & Dementia, 21(3), e14601. https://doi.org/10.1002/alz.14601
Zhu, Z., Shi, D., Guankai, P., Tan, Z., Shang, X., Hu, W., Liao, H., Zhang, X., Huang, Y., Yu, H., Meng, W., Wang, W., Ge, Z., Yang, X., & He, M. (2023). Retinal age gap as a predictive biomarker for mortality risk. British Journal of Ophthalmology, 107(4), 547–554. https://doi.org/10.1136/bjophthalmol-2021-319807
Nusinovici, S., Rim, T. H., Yu, M., Lee, G., Tham, Y. C., Cheung, N., Chong, C. C. Y., Soh, Z. D., Thakur, S., Lee, C. J., Sabanayagam, C., Lee, B. K., Park, S., Kim, S. S., Kim, H. C., Wong, T. Y., & Cheng, C. Y. (2022). Retinal photograph-based deep learning predicts biological age, and stratifies morbidity and mortality risk (RetiAGE). Age and Ageing, 51(4), afac065. https://doi.org/10.1093/ageing/afac065
Nusinovici, S., Rim, T. H., Li, H., Yu, M., Deshmukh, M., Quek, T. C., Lee, G., Chong, C. C. Y., Peng, Q., Xue, C. C., Zhu, Z., Chew, E. Y., Sabanayagam, C., Wong, T. Y., Tham, Y. C., & Cheng, C. Y. (2024). Application of a deep-learning marker for morbidity and mortality prediction derived from retinal photographs (RetiPhenoAge): A cohort development and validation study. The Lancet Healthy Longevity, 5(8), e514-e523. https://doi.org/10.1016/S2666-7568(24)00110-1
Yu, Z., Chen, R., Gui, P., Wang, W., Razzak, I., Alinejad-Rokny, H., Zeng, X., Shang, X., Zhang, L., Yang, X., Yu, H., Huang, W., Lu, H., van Wijngaarden, P., He, M., Zhu, Z., & Ge, Z. (2025). A cross-population study of retinal aging biomarkers with longitudinal pre-training and label distribution learning. npj Digital Medicine, 8, 331. https://doi.org/10.1038/s41746-025-00331-6
Zhou, Y., Chia, M. A., Wagner, S. K., Ayhan, M. S., Williamson, D. J., Struyven, R. R., Liu, T., Xu, M., Lozano, M. G., Woodward-Court, P., Kihara, Y., UK Biobank Eye & Vision Consortium, Altmann, A., Lee, A. Y., Topol, E. J., Denniston, A. K., Alexander, D. C., & Keane, P. A. (2023). A foundation model for generalizable disease detection from retinal images (RETFound). Nature, 622(7981), 156–163. https://doi.org/10.1038/s41586-023-06555-x
Qiu, J., Wu, J., Wei, H., Shi, P., Zhang, M., Sun, Y., Li, L., Liu, H., Liu, H., Hou, S., Zhao, Y., Shi, X., Xian, J., Qu, X., Zhu, S., Pan, L., Chen, X., Zhang, X., Jiang, S., . . . Yuan, W. (2024). Development and validation of a multimodal multitask vision foundation model for generalist ophthalmic artificial intelligence (VisionFM). NEJM AI, 1(12), AIoa2300221. https://doi.org/10.1056/AIoa2300221
Shi, D., Zhang, W., Chen, X., Liu, Y., Yang, J., Huang, S., Tham, Y. C., Zheng, Y., & He, M. (2024). EyeFound: A multimodal generalist foundation model for ophthalmic imaging. arXiv. https://doi.org/10.48550/arXiv.2405.11338
Engelmann, J., & Bernabeu, M. O. (2025). Training a high-performance retinal foundation model with half the data and 400 times less compute (RETFound-Green). Nature Communications, 16, 62123. https://doi.org/10.1038/s41467-025-62123-z
Hou, Q., Zhou, Y., Goh, J. H. L., Zou, K., Yew, S. M. E., Srinivasan, S., Wang, M., Lo, T. W. S., Lei, X., Wagner, S. K., Chia, M. A., Yang, G. D., Jiang, H., Ran, A. R., Santos, R., Somfai, G. M., Zhou, J. H., Chen, H., Chen, Q., . . . Tham, Y. C. (2025). Can a natural image-based foundation model outperform a retina-specific model in detecting ocular and systemic diseases? Head-to-head DINOv2 vs RETFound. Ophthalmology Science, 5(6), 100764. https://doi.org/10.1016/j.xops.2025.100764
Morano, J., Fazekas, B., Sükei, E., Fecso, R., Emre, T., Gumpinger, M., Faustmann, G., Oghbaie, M., Schmidt-Erfurth, U., & Bogunović, H. (2025). MIRAGE: Multimodal foundation model and benchmark for comprehensive retinal OCT image analysis. npj Digital Medicine, 8, 1852. https://doi.org/10.1038/s41746-025-01852-3
Judkiewicz, R., Berkowitz, E., Meisel, M., Michaeli, T., & Behar, J. A. (2026). Shifting the retinal foundation models paradigm from slices to volumes for OCT (V-JEPA). npj Digital Medicine, 9, 2496. https://doi.org/10.1038/s41746-026-02496-7
Farrah, T. E., Dhillon, B., Keane, P. A., Webb, D. J., & Dhaun, N. (2020). The eye, the kidney, and cardiovascular disease: Old concepts, better tools, and new horizons. Kidney International, 98(2), 323–342. https://doi.org/10.1016/j.kint.2020.01.039
Kellner, R. L., Harris, A., Ciulla, L., Guidoboni, G., Verticchio Vercellin, A., Oddone, F., Carnevale, C., Zaid, M., Antman, G., Kuvin, J. T., & Siesky, B. (2024). The eye as the window to the heart: Optical coherence tomography angiography biomarkers as indicators of cardiovascular disease. Journal of Clinical Medicine, 13(3), 829. https://doi.org/10.3390/jcm13030829
Ran, A. R., Tham, C. C., Chan, P. P., Cheng, C. Y., Tham, Y. C., Rim, T. H., & Cheung, C. Y. (2021). Deep learning in glaucoma with optical coherence tomography: A review. Eye, 35(1), 188–201. https://doi.org/10.1038/s41433-020-01191-5
Cheung, C. Y., Wong, W. L. E., Hilal, S., Kan, C. N., Gyanwali, B., Tham, Y. C., Schmetterer, L., Xu, D., Lee, M. L., Hsu, W., Venketasubramanian, N., Tan, B. Y., Wong, T. Y., & Chen, C. P. L. H. (2022). Deep-learning retinal vessel calibre measurements and risk of cognitive decline and dementia. Brain Communications, 4(4), fcac212. https://doi.org/10.1093/braincomms/fcac212
Lu, B., Li, Y., Xie, L., Chiu, K., Hao, X., Xu, J., Luo, J., & Sham, P. C. (2024). Computational retinal microvascular biomarkers from an OCTA image in clinical investigation. Biomedicines, 12(4), 868. https://doi.org/10.3390/biomedicines12040868
Cheung, C. Y., Mok, V., Foster, P. J., Trucco, E., Chen, C., & Wong, T. Y. (2021). Retinal imaging in Alzheimer’s disease. Journal of Neurology, Neurosurgery & Psychiatry, 92(9), 983–994. https://doi.org/10.1136/jnnp-2020-325347
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Adrianna Buż, Nicol Baran, Marta Bajkowska - Piterak, Jerzy Buszko, Damian Danilczuk, Wiktor Adamiec, Monika Szlachta-Gubernat, Przemysław Piterak, Michał Dyś
This work is licensed under a Creative Commons Attribution 4.0 International License.
All articles are published in open-access and licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0). Hence, authors retain copyright to the content of the articles.
CC BY 4.0 License allows content to be copied, adapted, displayed, distributed, re-published or otherwise re-used for any purpose including for adaptation and commercial use provided the content is attributed.
