Development of hybrid system for automatic diagnosis of diabetic retinopathy /
The Optic Nerve Head (ONH) and Vascular Intersection (VI) are important features in retina fundus image (RFI). The application of artificial intelligence has not received much attention in the diagnosis, prediction and monitoring of diabetic retinopathy (DR). The hybrid artificial intelligent system...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Kuala Lumpur :
Kulliyyah of Engineering, International Islamic University Malaysia,
2013
|
| Subjects: | |
| Online Access: | http://studentrepo.iium.edu.my/handle/123456789/4634 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The Optic Nerve Head (ONH) and Vascular Intersection (VI) are important features in retina fundus image (RFI). The application of artificial intelligence has not received much attention in the diagnosis, prediction and monitoring of diabetic retinopathy (DR). The hybrid artificial intelligent system includes image processing techniques and neural network trained with back propagation algorithm is proposed in this research work. Combined Cross Number points (CCN) technique which uses a 5×5 window embedded with Artificial Neural Network (ANN) technique has been proposed in the vasculature detection in order to detect the combination of bifurcation and crossover points in (RFI). On the other hand, three techniques are proposed in the ONH detection, namely simple thresholding technique and hybrid numerical differential approach as well as exponential histogram technique.Performance analysis of the system shows that ANN based technique for vascular intersection points detection achieves 100% accuracy on simulated images and a minimum of 92% accuracy on RFI obtained from DRIVE database. Thus, the simulated images have been used to train the artificial neural network (ANN) and on convergence the network is used to test (RFI) from DRIVE database. In the ONH detection, hybrid numerical differential and exponential histogram technique produces 90 % of accuracy as compared to the simple thresholding technique which only achieves 53% of accuracy. |
|---|---|
| Physical Description: | xiii, 113 leaves : ill. ; 30cm. |
| Bibliography: | Includes bibliographical references (leaves 81-85). |