A machine learning model was developed to predict the risk of diabetic retinopathy based on heavy metal exposure data. The model identified urinary Sb levels as the critical factor affecting the risk of DR, with age and Sb interaction being the most significant.
