A Dual-Stage Deep Lesion Segmentation Framework with Gradient and Boundary Optimization for Diabetic Retinopathy Retinal Images

Malathi  P; Dhinakaran D; Jeyalakshmi  S; Kavitha  P; Palpandi  S; Prabaharan  S

doi:10.35882/jeeemi.v8i4.1734

Malathi P Department of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, India https://orcid.org/0000-0003-3012-3629
Dhinakaran D Department of Computer Science and Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India https://orcid.org/0000-0002-3183-576X
Jeyalakshmi S Department of Artificial Intelligence and Data Science, P.S.R. Engineering College, Sivakasi, India https://orcid.org/0009-0004-6536-5083
Kavitha P Department of Computer Science and Engineering, R.M.K Engineering College, Chennai, India https://orcid.org/0000-0001-5109-1510
Palpandi S Department of Computer Science and Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India https://orcid.org/0000-0002-5293-6658
Prabaharan S Department of Computer Science and Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India https://orcid.org/0000-0002-8365-5322

DOI: https://doi.org/10.35882/jeeemi.v8i4.1734

Keywords: Diabetic Retinopathy, Lesion Boundaries, Gradient and Edge–Optimized Network, Boundary–Enhanced Segmentation Network, Deep Learning Models, Morphology, Edge Optimization

Abstract

Diabetic Retinopathy (DR) is a leading global cause of preventable blindness, and the early detection and precise segmentation of pathologies play a key role in clinical decisions in this disease. Conventional deep learning models, however, suffer from the problem of weak lesion boundaries, gradient inconsistency, and structural distortion in heterogeneous datasets. To overcome these drawbacks, we present a sophisticated systemthat combines GEONet (Gradient and Edge-Optimized Network) to achieve accurate gradient- and edge-aware segmentation, and BESNet (Boundary-Enhanced Segmentation Network) to introduce boundary refinement and structural preservation. This two-step strategy can guarantee the high-fidelity capture of subtle retinal lesions with anatomical consistency. Comparative analysis was carried out against state-of-the-art baselines. Experimental results on the EyePACS dataset demonstrate the effectiveness of the proposed framework. The integrated GEONet+BESNet architecture achieved a Dice Coefficient of 89.0%, Intersection-over-Union (IoU) of 87.0%, Boundary Accuracy of 89.0%, and a Hausdorff Distance of 5.1, outperforming all competing methods in terms of segmentation fidelity and boundary preservation. Structural consistency was further enhanced, attaining a Dice Similarity Coefficient (DSC) of 90.5% and a Structural Similarity Index Measure (SSIM) of 90.3%. From a clinical screening perspective, the proposed framework achieved a Precision of 92.8% and Specificity of 94.6%, indicating reliable lesion localization with a reduced rate of false-positive detections. These findings confirm that the synergistic integration of gradient-aware segmentation through GEONet and boundary-enhanced refinement through BESNet effectively preserves lesion morphology and improves segmentation robustness across heterogeneous retinal imaging conditions.

Downloads

Download data is not yet available.

References

R. Dahiya, N. Agarwal, S. Singh, D. Verma, and S. Gupta, “Diabetic Retinopathy Eye Disease Detection Using Machine Learning”, EAI Endorsed Trans IoT, vol. 10, Mar. 2024. doi:10.4108/eetiot.5349

M. Asif, F. Ur Rehman, Z. Rashid, A. Hussain, A. Mirza and W. S. Qureshi, "An Insight on the Timely Diagnosis of Diabetic Retinopathy Using Traditional and AI-Driven Approaches," in IEEE Access, vol. 13, pp. 116869-116886, 2025, doi: 10.1109/ACCESS.2025.3583647.

A. A. Khan, K. M. Ahmad, S. Shafiq, M. U. Akram, and J. Shao, “ATLASS: An anatomically-aware self-supervised learning framework for generalizable retinal disease detection,” IEEE Journal of Biomedical and Health Informatics, 2025.

B. Hassan, H. Raja, T. Hassan, M. U. Akram, H. Raja, A. A. Abdalrazaq, S. Yousefi, and N. Werghi, “A comprehensive review of artificial intelligence models for screening major retinal diseases,” Artificial Intelligence Review, vol. 57, no. 5, p. 111, 2024.

A. Tulsani, P. Kumar, and S. Pathan, “Automated segmentation of optic disc and optic cup for glaucoma assessment using improved UNETþþ architecture,” Biocybern. Biomed. Eng., vol. 41, no. 2, pp. 819–832, 2021, doi: 10.1016/j.bbe.2021.05.011.

W. T. Song, I. C. Lai, and Y. Z. Su, “A statistical robust glaucoma detection framework combining Retinex, CNN, and DOE using fundus images,” IEEE Access, vol. 9, pp. 103772–103783, 2021, doi: 10.1109/ACCESS.2021.3098032.

G. A. Saleh, N. M. Batouty, S. Haggag, A. Elnakib, F. Khalifa, F. Taher, M. A. Mohamed, R. Farag, H. Sandhu, A. Sewelam, and A. El-Baz, “The role of medical image modalities and ai in the early detection, diagnosis and grading of retinal diseases: a survey,” Bioengineering, vol. 9, no. 8, p. 366, 2022.

S. A. Khowaja, K. Dev, S. M. Anwar, and M. G. Linguraru, “Selffed: Self-supervised federated learning for data heterogeneity and label scarcity in medical images,” Expert Syst. Appl., vol. 261, p. 125493, 2025

S. Parsa and T. Khatibi, “Grading the severity of diabetic retinopathy using an ensemble of self-supervised pre-trained convolutional neural networks: Essp-cnns,” Multimedia Tools and Applications, vol. 83, pp. 89 837–89 870, 2024.

O. Kovalyk, J. Morales-Sanchez, R. Verd ´ u-Monedero, I. Sell ´ es-Navarro, ´ A. Palazon-Cabanes, and J.-L. Sancho-G ´ omez, “Papila: Dataset with ´ fundus images and clinical data of both eyes of the same patient for glaucoma assessment,” Scientific Data, vol. 9, no. 1, p. 291, 2022.

N. Mukherjee, S. Sengupta, M. Nadeem Ahmed, S. Irfan Yaqoob, M. Rashid Hussain and A. Taha Zamani, "Bi-Directional Hybrid Attention Feature Pyramid Network for Detecting Diabetic Macular Edema in Retinal Fundus Images," in IEEE Access, vol. 13, pp. 38726-38744, 2025, doi: 10.1109/ACCESS.2025.3545873.

D. Yi, P. Baltov, Y. Hua, S. Philip, and P. K. Sharma, “Compound scaling encoder-decoder (CoSED) network for diabetic retinopathy related bio-marker detection,” IEEE Journal of Biomedical and Health Informatics, early access, Sep. 11, 2023, doi: 10.1109/JBHI.2023.3313785.

L. Dai, L. Wu, H. Li, C. Cai, Q. Wu, H. Kong, R. Liu, X. Wang, X. Hou, Y. Liu, X. Long, Y. Wen, L. Lu, Y. Shen, Y. Chen, D. Shen, X. Yang, H. Zou, B. Sheng, et al., “A deep learning system for detecting diabetic retinopathy across the disease spectrum,” Nature Communications, vol. 12, no. 1, pp. 1–11, 2021, doi: 10.1038/s41467-021-23458-5.

A. Asia, C. Zhu, S. A. Althubiti, Y. Xiao, P. Ouyang, and M. A. A., “Detection of diabetic retinopathy in retinal fundus images using CNN classification models,” Electronics, vol. 11, no. 17, p. 2740, 2022, doi: 10.3390/electronics11172740.

A. Malhi, R. Grewal, and H. S. Pannu, “Detection and diabetic retinopathy grading using digital retinal images,” International Journal of Intelligent Robotics and Applications, vol. 7, pp. 426–458, 2023, doi: 10.1007/s41315-022-00269-5.

P. Modi and Y. Kumar, “Smart detection and diagnosis of diabetic retinopathy using bat based feature selection algorithm and deep forest technique,” Computers & Industrial Engineering, vol. 182, p. 109364, 2023, doi: 10.1016/j.cie.2023.109364.

K. Nazir, J. Kim and Y. -C. Byun, "Enhancing Early-Stage Diabetic Retinopathy Detection Using a Weighted Ensemble of Deep Neural Networks," in IEEE Access, vol. 12, pp. 113565-113579, 2024, doi: 10.1109/ACCESS.2024.3432867.

A. Zedadra, O. Zedadra, M. Yassine Salah-Salah and A. Guerrieri, "Graph-Aware Multimodal Deep Learning for Classification of Diabetic Retinopathy Images," in IEEE Access, vol. 13, pp. 74799-74810, 2025, doi: 10.1109/ACCESS.2025.3564529.

A. Jabbar, S. Naseem, J. Li, et al., “Deep transfer learning-based automated diabetic retinopathy detection using retinal fundus images in remote areas,” International Journal of Computational Intelligence Systems, vol. 17, p. 135, 2024, doi: 10.1007/s44196-024-00520-w.

I. Govindharaj, R. Rampriya, G. Michael, et al., “Capsule network-based deep learning for early and accurate diabetic retinopathy detection,” International Ophthalmology, vol. 45, p. 78, 2025, doi: 10.1007/s10792-024-03391-4.

Banupriya, V., Anusuya, S., "Strategy for Rapid Diabetic Retinopathy Exposure Based on Enhanced Feature Extraction Processing," Computers, Materials & Continua, 75(3), 5597–5613, 2023. doi:10.32604/cmc.2023.038696

Gour, N., & Khanna, P., "Automated glaucoma detection using GIST and pyramid histogram of oriented gradients (PHOG) descriptors," Pattern Recognition Letters, 137, 3-11, 2020. https://doi.org/10.1016/j.patrec.2019.04.004.

T. M. Devi, P. Karthikeyan, B. M. Kumar, and M. Manikandakumar, “Diabetic retinopathy detection via deep learning based dual features integrated classification model,” Technology and Health Care, 2025, doi: 10.1177/09287329241292939.

S. Rajeshwar, S. Thaplyal, A. M., and S. S. G., “Diabetic retinopathy detection using DL-based feature extraction and a hybrid attention-based stacking ensemble,” Advances in Public Health, vol. 2025, no. 1, p. 8863096, 2024, doi: 10.1155/adph/8863096.

R. Nandakumar, P. Saranya, V. Ponnusamy, S. Hazra, and A. Gupta, “Detection of diabetic retinopathy from retinal images using DenseNet models,” Computer Systems Science and Engineering, vol. 45, no. 1, pp. 279–292, 2023, doi: 10.32604/csse.2023.028703.

G. Sun, X. Wang, L. Xu, C. Li, W. Wang, Z. Yi, H. Luo, Y. Su, J. Zheng, Z. Li, Z. Chen, H. Zheng, and C. Chen, “Deep learning for the detection of multiple fundus diseases using ultra-widefield images,” Ophthalmology and therapy, vol. 12, no. 2, pp. 895–907, 2023.

J. Y. Choi, I. H. Ryu, J. K. Kim, I. S. Lee, and T. K. Yoo, “Development of a generative deep learning model to improve epiretinal membrane detection in fundus photography,” BMC Medical Informatics and Decision Making, vol. 24, no. 1, p. 25, 2024

F. Jeribi, T. Nazir, M. Nawaz, A. Javed, M. Alhameed, and A. Tahir, “Recognition of diabetic retinopathy and macular edema using deep learning,” Medical & Biological Engineering & Computing, pp. 1–15, 2024.

B. J. Bhatkalkar, D. R. Reddy, S. Prabhu, and S. V. Bhandary, “Improving the performance of convolutional neural network for the segmentation of optic disc in fundus images using attention gates and conditional random fields,” IEEE Access, vol. 8, pp. 29299–29310, 2020, doi: 10.1109/ACCESS.2020.2972318.

D. S. S. Raja, S. Kumarganesh, K. M. Sagayam, and H. Dang, “Diabetic Retinopathy Detection and Grading System Using Deep Learning Approach,” SAGE Open Medicine, 2026, doi: 10.1177/20552076251410982.

M. Alam, E. J. Zhao, C. K. Lam, and D. L. Rubin, “Segmentation-Assisted Fully Convolutional Neural Network Enhances Deep Learning Performance to Identify Proliferative Diabetic Retinopathy,” Journal of Clinical Medicine, vol. 12, no. 1, 2023, doi: 10.3390/jcm12010385.

S. Das, K. Kharbanda, M. Suchetha, R. Raman, and E. Dhas, “Deep Learning Architecture Based on Segmented Fundus Image Features for Classification of Diabetic Retinopathy,” Biomedical Signal Processing and Control, vol. 68, 2021, doi: 10.1016/j.bspc.2021.102600.

A. Sebastian, O. Elharrouss, S. Al-Maadeed, and N. Almaadeed, “A Survey on Diabetic Retinopathy Lesion Detection and Segmentation,” Applied Sciences, vol. 13, no. 8, 2023, doi: 10.3390/app13085111.

R. Raman, S. Srinivasan, S. Virmani, et al., “Fundus Photograph-Based Deep Learning Algorithms in Detecting Diabetic Retinopathy,” Eye, vol. 33, 2019, doi: 10.1038/s41433-018-0269-y.

A. Porwal et al., “IDRiD: Diabetic Retinopathy – Segmentation and Grading Challenge,” Medical Image Analysis, vol. 59, 2020, doi: 10.1016/j.media.2019.101561.

S. Dinesen, M. G. Schou, C. V. Hedegaard, et al., “A Deep Learning Segmentation Model for Detection of Active Proliferative Diabetic Retinopathy,” Ophthalmology and Therapy, vol. 14, pp. 1053–1063, 2025, doi: 10.1007/s40123-025-01127-w.

N. Sharma and P. Lalwani, “A Multi Model Deep Net with an Explainable AI Based Framework for Diabetic Retinopathy Segmentation and Classification,” Scientific Reports, vol. 15, 2025, doi: 10.1038/s41598-025-93376-9.

A. Rehman, G. Naijie, S. Ojo, et al., “FISM: Harnessing Deep Learning and Reinforcement Learning for Precision Detection of Microaneurysms and Retinal Exudates for Early Diabetic Retinopathy Diagnosis,” BioData Mining, vol. 18, 2025, doi: 10.1186/s13040-025-00485-2.

R. Nadda, J. Singh, and U. Shrivastava, “Automatic Diabetic Retinopathy Detection Using an Ensemble Learning Approach and Classifiers with Self-Adjusting Weights,” Soft Computing, vol. 29, 2025, doi: 10.1007/s00500-025-10773-y.

I. Govindharaj, A. Poongodai, G. Rajaram, et al., “Enhanced Diabetic Retinopathy Detection Using U-Shaped Network and Capsule Network-Driven Deep Learning,” MethodsX, vol. 14, 2025, doi: 10.1016/j.mex.2024.103052.