Prediction of systemic biomarkers from retinal photographs: development and validation of deep-learning algorithms

Abstract

Background The application of deep learning to retinal photographs has yielded promising results in predicting age, sex, blood pressure, and haematological parameters. However, the broader applicability of retinal photograph-based deep learning for predicting other systemic biomarkers and the generalisability of this approach to various populations remains unexplored. Methods With use of 236 257 retinal photographs from seven diverse Asian and European cohorts (two health screening centres in South Korea, the Beijing Eye Study, three cohorts in the Singapore Epidemiology of Eye Diseases study, and the UK Biobank), we evaluated the capacities of 47 deep-learning algorithms to predict 47 systemic biomarkers as outcome variables, including demographic factors (age and sex); body composition measurements; blood pressure; haematological parameters; lipid profiles; biochemical measures; biomarkers related to liver function, thyroid function, kidney function, and inflammation; and diabetes. The standard neural network architecture of VGG16 was adopted for model development. Findings In addition to previously reported systemic biomarkers, we showed quantification of body composition indices (muscle mass, height, and bodyweight) and creatinine from retinal photographs. Body muscle mass could be predicted with an R-2 of 0.52 (95% CI 0.51-0.53) in the internal test set, and of 0.33 (0.30-0.35) in one external test set with muscle mass measurement available. The R-2 value for the prediction of height was 0.42 (0.40-0.43), of bodyweight was 0.36 (0.34-0.37), and of creatinine was 0.38 (0.37-0.40) in the internal test set. However, the performances were poorer in external test sets (with the lowest performance in the European cohort), with R-2 values ranging between 0.08 and 0.28 for height, 0.04 and 0.19 for bodyweight, and 0.01 and 0.26 for creatinine. Of the 47 systemic biomarkers, 37 could not be predicted well from retinal photographs via deep learning (R-2=0.14 across all external test sets). Interpretation Our work provides new insights into the potential use of retinal photographs to predict systemic biomarkers, including body composition indices and serum creatinine, using deep learning in populations with a similar ethnic background. Further evaluations are warranted to validate these findings and evaluate the clinical utility of these algorithms.