Uddin, Minhaz

Abstract Background Diabetic retinopathy (DR) is the leading cause of preventable blindness among working-age adults worldwide, yet screening coverage remains inadequate, particularly in low-and middle-income countries. Automated deep learning systems offer potential to address the global shortage of expert graders, but most existing models lack lesion-level interpretability and are not aligned with established clinical referral frameworks. We developed and validated DRAGS (Diabetic Retinopathy Automated Grading System), a hybrid deep learning model that grades DR according to the UK Diabetic Eye Screening Programme (DESP) classification and provides lesion-level explainability. Methods We trained and validated a DenseNet-201-based convolutional neural network on 20,281 anonymised fundus images from two tertiary eye care institutions in Bangladesh. Images were graded by fellowship-trained retinal specialists using the UK DESP framework, resulting in 10 clinically interpretable classes that combine retinopathy grade (R0–R3) and maculopathy status (M0/M1). A companion dataset of 2,936 pixel-level lesion masks spanning nine pathological categories was used to train a parallel multi-label lesion-detection head. The dataset was partitioned 70:15:15 (patient-stratified). Performance was evaluated using macro-averaged AUROC (DeLong estimator), sensitivity, specificity, F1 score, quadratically weighted Cohen’s κ, and expected calibration error (ECE), with 95% CIs from 2000 bootstrap resamples. Grad-CAM spatial alignment with ground-truth lesion masks was assessed using Dice and IoU. This study follows the TRIPOD+AI reporting guidelines. Findings On the held-out test set (Component I: n = 3,044; Component II: n ≈ 440), DRAGS achieved class-wise precision, recall, and F1 scores ranging from 0·88 to 0·99 across all ten UK DESP grades, with advanced proliferative stages (R3–M0, R3–M1) consistently exceeding 0·95. Overall accuracy was approximately 91·1% and quadratically weighted Cohen’s κ was approximately 0·90. For referable versus non-referable DR, sensitivity was 90·7% and specificity was 91·9%. The companion lesion-detection head achieved macro-averaged sensitivity of 93·9%, specificity of 99·5%, and AUC of 0·997 across nine lesion classes; seven of nine classes achieved AUC = 1·00. Grad-CAM activations showed progressive spatial shift from diffuse (normal) to lesion-dense peripheral patterns (proliferative DR), with maximal agreement for microaneurysms and exudates. Mean inference time was 110–160 ms per image. Interpretation DRAGS demonstrates high diagnostic accuracy for nine-class UK DESP-aligned DR grading, with clinically interpretable lesion-level explainability on a large real-world LMIC dataset. External validation and prospective clinical evaluation are warranted before deployment. Funding The present study received no funding.