Related Experiment Video
Updated: Dec 12, 2025
07:11
Assessing Early Stage Open-Angle Glaucoma in Patients by Isolated-Check Visual Evoked Potential
Published on: May 25, 2020
6.7K
Controlled differential evolution based detection of neovascularization on optic disc using support vector machine
Birendra Biswal1, Geetha Pavani P1, Pradyut K Biswal2
1Department of ECE, Gayatri Vidya Parishad College of Engineering (A), Visakhapatnam, Andhra Pradesh, India.
Biomedizinische Technik. Biomedical Engineering
|August 11, 2020
Summary
Early detection of Neovascularization on Disc (NVD) in Diabetic Retinopathy (DR) is crucial for preventing blindness. This study introduces an automated method using Controlled Differential Evolution and Support Vector Machine for accurate NVD identification.
Area of Science:
- Ophthalmology
- Medical Imaging
- Computational Biology
Background:
- Diabetic Retinopathy (DR) can lead to blindness due to Neovascularization on Disc (NVD).
- Accurate and early detection of NVD is critical for timely intervention and vision preservation.
Purpose of the Study:
- To develop an automated system for detecting Neovascularization on Disc (NVD) in fundus images.
- To improve the accuracy and efficiency of NVD diagnosis in Diabetic Retinopathy.
Main Methods:
- An automated optic disc detection using Controlled Differential Evolution (CDE) algorithm.
- Region of Interest (ROI) extraction around the optic disc for NVD analysis.
- Blood vessel segmentation using multi-scale Gabor filtering.
- Extraction of morphological, textural, and statistical features.
- Classification using Support Vector Machine (SVM) after feature selection (Sequential Backward Feature - SBF).
Main Results:
- The system achieved high performance on 205 fundus images (99 NVD positive, 106 NVD negative).
- Average accuracy of 98.75%, specificity of 100%, sensitivity of 97.8%, and AUC of 100% were attained.
- Feature dimension was reduced from 16 to 13 optimal features.
Conclusions:
- The proposed automated method demonstrates high accuracy and reliability for NVD detection.
- This approach can aid in the early diagnosis and management of Diabetic Retinopathy.
- The system shows potential for clinical application in ophthalmology.
