Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study introduces PRISM-DR, a specialized pipeline for detecting diabetic retinopathy lesions by employing individual single-class detectors for each lesion type, addressing the limitations of traditional multi-class models that often overlook rare lesions. The method incorporates region of interest cropping, YOLO-based detection, and inter-lesion suppression to enhance detection accuracy, achieving a test mAP50 of 0.527 and an F1 score of 0.529, with the best performance on hard exudates. While the results are modest due to the challenges of the task and the training dataset, the approach demonstrates a promising alternative for improving detection rates in clinical settings.
Treating each diabetic retinopathy lesion as a separate detection problem significantly enhances detection accuracy, especially for rare lesions.
Diabetic retinopathy is a leading cause of preventable blindness; its early lesions are small, low contrast, and easily missed in manual screening. Most automated detectors handle the four non-proliferative DR lesions: microaneurysms, hemorrhages, hard exudates, and soft exudates, with a single multi-class model, even though these lesions differ sharply in size, color, morphology, and prevalence, so a shared model favors common, easy classes over rare, difficult ones. We present PRISM-DR, a lesion-specific pipeline that trains one single-class detector per lesion, each with its own configuration. From a raw fundus image, the pipeline applies region of interest cropping, fundus-specific preprocessing, four parallel YOLO detectors, tiling, per-lesion ensembling of five cross-validation folds, and an inter-lesion suppression step that resolves overlaps by physical lesion size and clinical priority rather than confidence. Per lesion, the best of five YOLO generations is selected, and augmentation is tuned by Bayesian optimization. Trained on IDRiD with stratified five-fold cross-validation, the system reaches a test mAP50 of 0.527 and F1 of 0.529, highest AP50 on hard exudates with 0.561. Without fine-tuning, the models transfer well where the imaging scale is close to IDRiD and degrade as field of view and resolution depart. These modest absolute results reflect a small single-source training set and a difficult task; however, treating each lesion as a separate detection problem is a practical alternative to a single multi-class model.