Related Experiment Video
Updated: Mar 5, 2026
Author Spotlight: Understanding Age-Related Macular Degeneration Pathophysiology with QAF Workflow
Published on: May 26, 2023
Automated diabetic macular edema (DME) grading system using DWT, DCT Features and maculopathy index
U Rajendra Acharya1, Muthu Rama Krishnan Mookiah2, Joel E W Koh2
1Department of Electronics and Computer Engineering, Ngee Ann Polytechnic, Singapore 599489, Singapore; Department of Biomedical Engineering, School of Science and Technology, SIM University, Singapore 599491, Singapore; Department of Biomedical Engineering, Faculty of Engineering, University of Malaya, 50603, Malaysia.
This study presents an automated system for detecting diabetic macular edema (DME) using Radon, wavelet, and cosine transforms. The hybrid approach achieves high accuracy, offering a potential tool for early diabetic eye screening.
Area of Science:
- Ophthalmology
- Medical Imaging
- Computer Science
Background:
- Diabetic macular edema (DME) is a vision-threatening complication of diabetes mellitus (DM).
- Clinical diagnosis of DME from fundus images is time-consuming and labor-intensive.
- Early detection and treatment are crucial to prevent vision loss in diabetic patients.
Purpose of the Study:
- To develop and validate a hybrid automated system for the detection and grading of diabetic macular edema (DME).
- To evaluate the efficacy of combining Radon Transform (RT), Discrete Wavelet Transform (DWT), and Discrete Cosine Transform (DCT) for DME classification.
- To formulate a maculopathy index for distinct discrimination of DME severity.
Main Methods:
- A hybrid image processing technique combining Radon Transform (RT), Discrete Wavelet Transform (DWT), and Discrete Cosine Transform (DCT) was employed.
- Fundus images were processed through RT to generate sinograms, followed by DWT for feature extraction.
- DCT was applied to approximate coefficients, and features were analyzed using Locality Sensitive Discriminant Analysis (LSDA) and supervised classifiers.
Main Results:
- The automated system achieved high classification accuracies of 100% and 97.01% on private and public (MESSIDOR) datasets, respectively.
- The system utilized a minimal number of significant features (two and seven) for classification.
- A novel maculopathy index was formulated, enabling clear discrimination between three DME severity groups using a single integer.
Conclusions:
- The proposed hybrid system demonstrates high accuracy and efficiency for automated diabetic macular edema detection.
- This system shows significant potential as a non-invasive, automated eye screening tool for diabetic individuals.
- The developed maculopathy index offers a simplified method for assessing DME severity.
