EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Deep learning-based analysis of retinal images for estimation of cardiovascular disease risk: the multi-ethnic study of atherosclerosis (MESA)

Ching Yu Cheng, Sahil Thakur, Simon Nusinovici, Jose David Vargas Quiros, Marco Yu, Dongjin Nam, Bart Liefers, Craig Johnson, Aaron Lee, Chan Joo Lee, Charumathi Sabanayagam, Daniel Bos und 14 weitere

American Journal of Preventive Cardiology · 2026-10 · 102573

DOI: 10.1016/j.ajpc.2026.102573 ↗

Kurz gefasst

Fachartikel in der Zeitschrift American Journal of Preventive Cardiology, 2026-10. Verfasst von Ching Yu Cheng und 25 weiteren Autor:innen. Die Arbeit nennt 45 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Medical Research Council.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Deep learning retinal scores for cardiovascular risk have been validated only in East Asian and European-ancestry cohorts. We evaluated one in a US multi-ethnic cohort with adjudicated events and measured coronary artery calcium (CAC).

Methods

RetiCAC is a deep learning score estimating the probability of CAC presence from retinal photographs. In Multi-Ethnic Study of Atherosclerosis (MESA) participants without baseline atherosclerotic cardiovascular disease (ASCVD), RetiCAC was categorized by quartiles (Q1 low, Q2–Q3 moderate, Q4 high) and by fixed cut points (30.9 and 69.0) derived against measured CAC without reference to outcomes, with pre-specified hypotheses on observed 10-year risk. A retinal 10-year ASCVD risk estimator combining RetiCAC with age and sex was applied after baseline-hazard recalibration. We assessed stratification (adjusted Cox models), discrimination (c-index), reclassification (continuous NRI), calibration overall and by race/ethnicity (modified Hosmer–Lemeshow [HL] test), and residual association beyond CAC.

Results

Among 4,890 participants (mean age 63±10 years; 52.2% female; White, Black, Hispanic, and Chinese; follow-up 9.0±2.3 years), 9.7% had an event. Among those not high risk by Pooled Cohort Equations (PCE; n=3,469), high-risk RetiCAC had adjusted HR 4.17 (95% CI 2.24–7.76) versus low, comparable to CAC >100 (HR 3.85, 95% CI 2.70–5.50); c-indices were 0.708 (0.675–0.742) and 0.731 (0.699–0.763; p=0.098). Fixed cut points separated groups with observed 10-year risks of 2.3%, 6.6%, and 15.7% (c-index 0.712), meeting both hypotheses. Added to PCE, RetiCAC raised the c-index by 0.015 (95% CI 0.008–0.024; NRI 0.26); CAC by 0.046 (0.034–0.061; NRI 0.54), with similar gains beyond PREVENT. Residual RetiCAC not explained by CAC remained associated with events (HR 1.33 per SD, 95% CI 1.13–1.56). The retinal estimator was calibrated after recalibration (intercept −0.46, slope 0.74; HL p=0.044) and in all four race/ethnicity groups (HL p=0.14–0.96), with a c-index of 0.705 (0.684–0.727), comparable to PCE.

Conclusions

In a US multi-ethnic cohort, RetiCAC stratified ASCVD risk among individuals at low or moderate risk by PCE, with fixed thresholds that identified a group at high observed 10-year risk and carried prognostic information beyond CAC itself. Retinal photographs, already acquired at scale in eye care, can support cardiovascular risk stratification where CT calcium scoring is unavailable, pending prospective evaluation of these thresholds.

Förderung

  • National Medical Research Council

Zitieren

APA

Cheng, C. Y., Thakur, S., Nusinovici, S., Quiros, J. D. V., Yu, M., Nam, D., Liefers, B., Johnson, C., Lee, A., Lee, C. J., Sabanayagam, C., Bos, D., Fernandez, I. R., Lahousse, L., Nasir, K., Bittencourt, M. S., Arsanjani, R., Lester, S. J., Rim, T. H., …, & Kavousi, M. (2026). Deep learning-based analysis of retinal images for estimation of cardiovascular disease risk: the multi-ethnic study of atherosclerosis (MESA). American Journal of Preventive Cardiology, 102573. https://doi.org/10.1016/j.ajpc.2026.102573

Vancouver

Cheng CY, Thakur S, Nusinovici S, Quiros JDV, Yu M, Nam D, et al. Deep learning-based analysis of retinal images for estimation of cardiovascular disease risk: the multi-ethnic study of atherosclerosis (MESA). American Journal of Preventive Cardiology. 2026:102573. doi:10.1016/j.ajpc.2026.102573

BibTeX
@article{Cheng2026,
  author = {Cheng, Ching Yu and Thakur, Sahil and Nusinovici, Simon and Quiros, Jose David Vargas and Yu, Marco and Nam, Dongjin and Liefers, Bart and Johnson, Craig and Lee, Aaron and Lee, Chan Joo and Sabanayagam, Charumathi and Bos, Daniel and Fernandez, Idalys R and Lahousse, Lies and Nasir, Khurram and Bittencourt, Marcio Sommer and Arsanjani, Reza and Lester, Steven J and Rim, Tyler Hyungtaek and Klaver, Caroline C.W. and Park, Sungha and Sperling, Laurence S. and Shapiro, Michael D and McClelland, Robyn L. and Maron, David J. and Kavousi, Maryam},
  title = {Deep learning-based analysis of retinal images for estimation of cardiovascular disease risk: the multi-ethnic study of atherosclerosis (MESA)},
  journal = {American Journal of Preventive Cardiology},
  year = {2026},
  pages = {102573},
  doi = {10.1016/j.ajpc.2026.102573}
}
Kostenlos starten →