Appropriate identification of age-related macular degeneration using OCT images

• Published in: Computer methods in biomechanics and biomedical engineering. Vol. 9; no. 2; pp. 146 - 156
• Main Authors: Hani, Marwa, Ben Slama, Amine, Zghal, Imen, Trabelsi, Hedi
• Format: Journal Article
• Language: English
• Published: Taylor & Francis 04.03.2021
• Subjects: Convolutional Neural Networks (CNNs); macular region segmentation; neural networks; Optical coherence tomography; support vector machines; watersheds
• ISSN: 2168-1163; 2168-1171
• DOI: 10.1080/21681163.2020.1827041

This paper presents an automatic approach for macular degeneration diagnostic by adding pertinent features in the evaluation process. In macular abnormality identification, the manual assessment of retina in OCT scan is an error prone task, especially in the preliminary behaviour levels. The efficiency of this presence is related to experience and the attention of ophthalmologists. Thereby, different techniques of retinal-image analysis help the option of obtaining a consistent, independent and overhead all rapid diagnosis to identify macular degeneration behaviour. We propose a computerised technique based on morphological parameters and a classification scheme to categorise subjects into two topics: normal and age related to macular degeneration (AMD). The obtained results prove that the use of a support vector machine (SVM) method is an efficient way to improve recognition for fast and accurate ophthalmology diagnostic.