Computer Aided Cataract Diagnosis using Fundus Retinal Images

Loading...
Thumbnail Image

Date

item.page.authors

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Cataract is the cloudiness present in the eye lens due to the denaturation of active protein cells. Cataract affects the quality of life and thereby impacting daily routine activities. Cataract may cause blindness if it is not detected at an earlier stage. Early detection and intervention could prevent vision loss and slow the development of cataracts. The computer-aided cataract diagnosis (CACD) method utilizing fundus retinal images is necessary to diagnose a large-screen population. There are three CACD approaches suggested in this thesis. Out of the three CACD methods that are currently being discussed, one way focuses on enhancing diagnostic performance based on accuracy, while the other two methods concentrate on enhancing robustness. A noise level estimation (NLE) method is also suggested in addition to the CACD methods. newlineInitially, a CACD method using transfer learning is proposed to detect various stages of the cataract such as normal, mild, moderate, and severe from the fundus retinal images. The fundus images with and without cataract are collected from the various open-access datasets and then labeled into four classes with the help of ophthalmologic experts. The proposed method uses the pre-trained deep neural network (DNN) for transfer learning to carry out automatic cataract classification. The pre-trained DNN model is used for the feature extraction, and the extracted features are then given to the support vector machine (SVM) classifier for accurate diagnosis. It is observed that the quality of images plays a vital role in effective clinical cataract diagnosis. An image quality selection module is thus incorporated into the proposed CACD method to ensure the required fundus image quality for diagnosis.

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced