Figure 1
From: Age and sex affect deep learning prediction of cardiometabolic risk factors from retinal images
Model performance on predicting continuous cardiometabolic risk factors in the test set on person level for the regression task. The coefficient of determination is plotted for every risk factor, along with the 95% confidence interval obtained via a bootstrapping methodology. Results for a linear regression using age and sex on person level on the test set is added for every cardiometabolic risk factor, except age, with a red dot. Risk factors included in the exploration of the impact of age and sex on prediction performance have a coefficient of determination higher than 0.20 and are indicated in dark green.
