Cross Feature Fusion of Fundus Image and Generated Lesion Map for Referable Diabetic Retinopathy Classification
This work addresses early detection of diabetic retinopathy to prevent blindness, offering a potential clinical tool, but it is incremental as it builds on existing segmentation and attention techniques.
The paper tackled referable diabetic retinopathy classification by developing a cross-learning method that uses fundus images and generated lesion maps as complementary inputs with a cross-attention mechanism, achieving 94.6% accuracy and surpassing state-of-the-art methods by 4.4% on public datasets.
Diabetic Retinopathy (DR) is a primary cause of blindness, necessitating early detection and diagnosis. This paper focuses on referable DR classification to enhance the applicability of the proposed method in clinical practice. We develop an advanced cross-learning DR classification method leveraging transfer learning and cross-attention mechanisms. The proposed method employs the Swin U-Net architecture to segment lesion maps from DR fundus images. The Swin U-Net segmentation model, enriched with DR lesion insights, is transferred to generate a lesion map. Both the fundus image and its segmented lesion map are used as complementary inputs for the classification model. A cross-attention mechanism is deployed to improve the model's ability to capture fine-grained details from the input pairs. Our experiments, utilizing two public datasets, FGADR and EyePACS, demonstrate a superior accuracy of 94.6%, surpassing current state-of-the-art methods by 4.4%. To this end, we aim for the proposed method to be seamlessly integrated into clinical workflows, enhancing accuracy and efficiency in identifying referable DR.