Abstract

Diabetic Retinopathy (DR) is a leading cause of preventable blindness. Early detection can identify retinal lesions such as microaneurysms, hemorrhages, and exudates. Traditional diagnosis is often slow and inconsistent to address this issue, this study introduces a Lesion-Aware DR Detection framework. It combines U-Net for pixel-level lesion segmentation with a Hybrid Attention Network (HAN) for better feature refinement. The U-Net architecture gathers information at multiple scales. The hybrid attention mechanism uses both spatial and channel attention. This allows the model to focus on important lesion areas and ignore unimportant background features. This combined approach improves the detection of small lesions and enhances the learning process for DR grading. The suggested method outperforms baseline U-Net and traditional CNN models in lesion segmentation accuracy, sensitivity, and overall DR classification, according to experimental results on standard DR datasets. The diagnosis of diabetic retinopathy was greatly aided by lesion-aware techniques using U-Net for pixel-level lesion segmentation and Hybrid Attention Network (HAN) for improved feature refinement. The framework presents a dependable assistive tool for automated, early DR screening and improves interpretability.

References

• Ronneberger, O., Fischer, P., Brox, T. (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation. In: Navab, N., Hornegger, J., Wells, W., Frangi, A. (eds) Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. MICCAI 2015. Lecture Notes in Computer Science(), vol 9351. Springer, Cham. https://doi.org/10.1007/978-3-319-24574-4_28
• Saini, Lalit. “Role of Artificial Intelligence in Medical Devices.” IJRASET, vol. 12, no. 12, IJRASET Research and Development Corp, 2024, pp. 1363–1367.
• Fu L, Li S, Sun Y, Mu Y, Hu T, Gong H. Lightweight-Convolutional Neural Network for Apple Leaf Disease Identification. Front Plant Sci. 2022 May 24;13:831219. doi: 10.3389/fpls.2022.831219. PMID: 35685005; PMCID: PMC9171387.
• Abolfazl younesi , mohsen ansari , mohammadamin fazli , alireza ejlali , muhammad shafique , (Senior Member, IEEE), and örg henkel A Comprehensive Survey of Convolutions in Deep Learning: Applications, Challenges, and Future Trends, (Fellow, IEEE)
• Harry Pratt, Frans Coenen, Deborah M. Broadbent, Simon P. Harding, Yalin Zheng, Convolutional Neural Networks for Diabetic Retinopathy,Procedia Computer Science,Volume 90,2016, Pages 200-205, ISSN 1877-0509.
• Wang, Zhe, Yanxin Yin, Jianping Shi, Wei Fang, Hongsheng Li, and Xiaogang Wang. "Zoom-in-net: Deep mining lesions for diabetic retinopathy detection." In International conference on medical image computing and computer-assisted intervention, pp. 267-275. Cham: Springer International Publishing, 2017.
• Asia, A.-O.; Zhu, C.-Z.; Althubiti, S.A.; Al-Alimi, D.; Xiao, Y.-L.; Ouyang, P.-B.; Al-Qaness, M.A.A. Detection of Diabetic Retinopathy in Retinal Fundus Images Using CNN Classification Models. Electronics 2022, 11, 2740. https://doi.org/10.3390/electronics11172740
• Lam C, Yi D, Guo M, Lindsey T. Automated Detection of Diabetic Retinopathy using Deep Learning. AMIA Jt Summits Transl Sci Proc. 2018 May 18;2017:147-155. PMID: 29888061; PMCID: PMC5961805.
• Oktay, O., Schlemper, J., Folgoc, L. L., et al. (2018). Attention U-Net: Learning Where to Look for the Pancreas. Medical Image Analysis, vol. 53, pp. 1–14.
• Mohammad Aknan, Salmoli Chandra, Shivani Gupta, Juan Carlos Cotrina Aliaga, and Smaranika Mohapatra. 2023. A Diabetic Retinopathy Classification and Analysis Towards the Development of a Computer-Aided Medical Decision Support System. SN Comput. Sci. 4, 6 (Oct 2023). https://doi.org/10.1007/s42979-023-02178-1
• Chaturvedi Kapil, Vijay Bhandari, Anil Kothari, Ritu Shrivastava, and Rishabh Tiwari. "Enhanced Diabetic Retinopathy Detection Mechanism Through Ensemble Learning Techniques." In International Conference on Data Science and Big Data Analysis, pp. 17-37. Singapore: Springer Nature Singapore, 2024.
• Chaturvedi Kapil, Vijay Bhandari, Ritu Shrivastava, and Rishabh Tiwari. "A Recent Review on Diabetic Retinopathy Detection (DRD) Approaches." In International Conference on Advanced Computing and Applications, pp. 559-574. Singapore: Springer Nature Singapore, 2024.
• K V Spoorthi; B S Rekha, “Diabetic Retinopathy Prediction using Deep learning”.
• E. T. D. R. S. R. Group. Grading diabetic retinopa thy from stereoscopic color fundus photographs- an extension of the modified Airlie House classification. Ophthalmology 1991;98(5):786–806.