A deep-learning system for the assessment of CVD risk via the measurement of retinal-vessel calibre

A deep-learning system for the assessment of CVD risk via the measurement of retinal-vessel calibre

Share this post

Choose a social network to share with, or copy the shortened URL to share elsewhere

This is a representation of how your post may appear on social media. The actual post will vary between social networks

Cardiovascular diseases (CVD) is the most common cause of death worldwide. While traditional risk factors, such as hypertension, hyperlipidemia and cigarette smoking, allow physicians to identify, monitor and treat patients at high-risk of CVD, a substantial proportion of CVD morbidity is not explained by these traditional risk factors. As a result, there is increasing interest in finding additional biomarkers for CVD risk stratification.

The retinal blood vessels, which are accessible to direct non-invasive visualization, share similar anatomical and physiological characteristics with microcirculation in the body. For over three decades, many researchers have sought to measure retinal blood vessel changes from photographs of the retina using imaging software, typically with human assessors. In our recent study in Nature Biomedical Engineering, we developed an artificial intelligence (AI) deep learning convolutional neural network (CNN) to measure the calibre of retinal blood vessels. We were able to test this AI system in diverse multi-ethnic, multi-country datasets in more than 60,000 retinal images. We found a high correlation of the AI system with validated human measurements, and demonstrated comparable (or stronger) relationships of AI-measured retinal vessel calibre with classic CVD risk factors (e.g., blood pressure). We were also able to demonstrate that AI-measured retinal vessel calibre could be predictive of incident CVD events.

So what is the next step? We are in contact with other collaborative groups to test the AI system, although there are challenges in data sharing in the current environment. We hope to create a “Retinal vessel CVD risk score” based on the AI system and compare this with traditional CVD risk equations. We are excited that advancement in AI in medical imaging has now allowed us to significantly improve our ability to study the human microcirculation via the retina vessels. Further evaluation of the utility of the AI system will allow translation to improve clinical management of CVD.


  1. Wong, T.Y. & Mitchell, P. Hypertensive retinopathy. N Engl J Med 351, 2310-2317 (2004).
  2. Liew, G., Wang, J.J., Mitchell, P. & Wong, T.Y. Retinal vascular imaging: a new tool in microvascular disease research. Circ.Cardiovasc.Imaging 1, 156-161 (2008).
  3. Cheung, C.Y., Ikram, M.K., Chen, C. & Wong, T.Y. Imaging retina to study dementia and stroke. Prog Retin Eye Res 57, 89-107 (2017).
  4. Cheung, C.Y., Ikram, M.K., Sabanayagam, C. & Wong, T.Y. Retinal microvasculature as a model to study the manifestations of hypertension. Hypertension 60, 1094-1103 (2012).
  5. Wang, S.B., et al. A spectrum of retinal vasculature measures and coronary artery disease. Atherosclerosis 268, 215-224 (2018).
  6. Hubbard, L.D., et al. Methods for evaluation of retinal microvascular abnormalities associated with hypertension/sclerosis in the Atherosclerosis Risk in Communities Study. Ophthalmology 106, 2269-2280 (1999).
  7. Wong, T.Y., et al. Computer-assisted measurement of retinal vessel diameters in the Beaver Dam Eye Study: methodology, correlation between eyes, and effect of refractive errors. Ophthalmology 111, 1183-1190 (2004).
  8. McGrory, S., et al. Towards Standardization of Quantitative Retinal Vascular Parameters: Comparison of SIVA and VAMPIRE Measurements in the Lothian Birth Cohort 1936. Transl Vis Sci Technol 7, 12 (2018).
  9. Cheung, C.Y., et al. A new method to measure peripheral retinal vascular caliber over an extended area. Microcirculation 17, 495-503 (2010).
  10. Cheung, C.Y., et al. Retinal vascular tortuosity, blood pressure, and cardiovascular risk factors. Ophthalmology 118, 812-818 (2011).
  11. Cheung, C.Y., et al. Quantitative and Qualitative Retinal Microvascular Characteristics and Blood Pressure J Hypertens. (2011).
  12. Cheung, C.Y., et al. Retinal vascular fractal dimension and its relationship with cardiovascular and ocular risk factors. Am J Ophthalmol 154, 663-674 e661 (2012).
  13. Cheung, C.Y., et al. Retinal microvascular changes and risk of stroke: the Singapore Malay Eye Study. Stroke 44, 2402-2408 (2013).
  14. Ho, H., et al. Retinopathy Signs Improved Prediction and Reclassification of Cardiovascular Disease Risk in Diabetes: A prospective cohort study. Sci Rep 7, 41492 (2017).
  15. Yip, W., et al. Joint Effect of Early Microvascular Damage in the Eye &Kidney on Risk of Cardiovascular Events. Sci Rep 6, 27442 (2016).
  16. Arnould, L., et al. Association between the retinal vascular network with Singapore "I" Vessel Assessment (SIVA) software, cardiovascular history and risk factors in the elderly: The Montrachet study, population-based study. PLoS One 13, e0194694 (2018).

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in