Automated classification of diabetic retinopathy through reliable feature selection
- PDF / 1,490,297 Bytes
- 19 Pages / 595.276 x 790.866 pts Page_size
- 25 Downloads / 208 Views
SCIENTIFIC PAPER
Automated classification of diabetic retinopathy through reliable feature selection S. Gayathri1 · Varun P. Gopi1 · P. Palanisamy1 Received: 2 November 2019 / Accepted: 23 June 2020 © Australasian College of Physical Scientists and Engineers in Medicine 2020
Abstract Diabetic retinopathy (DR) is a complication of diabetes mellitus that damages the blood vessels in the retina. DR is considered a serious vision-threatening impediment that most diabetic subjects are at risk of developing. Effective automatic detection of DR is challenging. Feature extraction plays an important role in the effective classification of disease. Here we focus on a feature extraction technique that combines two feature extractors, speeded up robust features and binary robust invariant scalable keypoints, to extract the relevant features from retinal fundus images. The selection of top-ranked features using the MR-MR (maximum relevance-minimum redundancy) feature selection and ranking method enhances the efficiency of classification. The system is evaluated across various classifiers, such as support vector machine, Adaboost, Naive Bayes, Random Forest, and multi-layer perception (MLP) when giving input image features extracted from standard datasets (IDRiD, MESSIDOR, and DIARETDB0). The performances of the classifiers were analyzed by comparing their specificity, precision, recall, false positive rate, and accuracy values. We found that when the proposed feature extraction and selection technique is used together with MLP outperforms all the other classifiers for all datasets in binary and multiclass classification. Keywords DR detection · Retinal fundus images · SURF · BRISK · Feature selection and ranking · MR-MR method · 10Fold cross validation
Introduction Diabetic retinopathy (DR) mainly occurs in people who have a history of diabetes mellites. DR is a result of gradual damage to the blood vessels in the retina; where these tiny blood vessels eventually leak blood and other fluids. DR is considered a serious vision-threatening complication of diabetes. DR is a clearly defined marker of coronary diseases[1]. The significance of early detection of DR is that it might help to reduce the risk of coronary diseases and can be used as a biomarker for many chronic diseases. However, early detection of DR requires efficient algorithms that can extract the most valuable features from the image for classifying images as either DR or normal fundus. The main stages involved in DR can be categorized as non-proliferative DR (NPDR)
* Varun P. Gopi [email protected] S. Gayathri [email protected] 1
National Institute of Technology, Trichy, Tamilnadu, India
and proliferative DR (PDR). At the primary stage (NDPR), many symptoms might develop in the retina due to the DR. PDR refers to the advanced stage of DR. The stages in DR are broadly categorized as mild NPDR, moderate NPDR, severe NPDR, and PDR[2]. This category is required when DR severity grading is considered. The structure of a normal retina is shown in Fig 1a
Data Loading...
