Machine learning algorithms developed to predict improvement in patients with mild nonproliferative diabetic retinopathy using retinal imaging features were more accurate compared with algorithms using systemic measurements.
“Our results indicate that the results of all feature families provided the best predictive outcome for DR improvement in mild NPDR patients. Features based on retinal imaging showed a higher predictive value than the demographic features and systemic measurements. These results are consistent with those of prior works,” Dimitrios Damopoulos, PhD, said at the
Uncategorized