|
|
 |
| |
|
|
|
Automated Detection of Diabetic Retinopathy from Fundus Images Using Classical Image Processing |
|
|
|
PP: 1141-1158 |
|
|
doi:10.18576/amis/200502
|
|
|
|
Author(s) |
|
|
|
Haitham Issa,
Hani Attar,
Jafar Ababneh,
Zakaria Che Muda,
Ismail A. M. Elhaty,
Ahmed Solyman,
Ramy M. Bahy,
|
|
|
|
Abstract |
|
|
| Diabetic retinopathy (DR) is a leading cause of visual impairment and blindness in patients with diabetes, making early and accurate screening essential for timely clinical intervention. This paper presents an automated framework for DR detection and severity assessment using retinal fundus images, classical image processing techniques, and lesion-based feature extraction. The proposed method consists of five main stages: image preprocessing, blood-vessel detection, optic-disc removal, microaneurysm and hemorrhage detection, and rule-based severity classification. In the preprocessing stage, fundus images are converted to grayscale, enhanced using contrast-limited adaptive histogram equalization (CLAHE), and processed with circular masking to remove non-retinal background regions. The feature extraction stage identifies retinal blood vessels and detects key DR-related lesions, particularly microaneurysms and hemorrhages. The extracted lesion features are then used to classify the severity of DR through an interpretable rule-based decision process. The framework was evaluated using 1,000 fundus images representing different DR severity levels and achieved an overall classification accuracy of 93.75%. The results indicate that the proposed approach offers an effective, transparent, and low-complexity solution for automated DR screening. Therefore, it can provide practical support for ophthalmologists and healthcare systems, particularly in large-scale screening programs and resource-constrained clinical environments. This work also contributes to the United Nations Sustainable Development Goal 3 (Good Health and Well-Being) by supporting early, accessible, and cost-effective screening for diabetic eye disease. |
|
|
|
|
|
|
|
