Simulation of diabetic retinopathy utilizing convolutional neural networks

Rajarajeswari, P, Moorthy, J and Beg, OA ORCID: https://orcid.org/0000-0001-5925-6711 2021, 'Simulation of diabetic retinopathy utilizing convolutional neural networks' , Journal of Mechanics in Medicine and Biology . (In Press)

[img] PDF - Accepted Version
Restricted to Repository staff only

Download (1MB) | Request a copy

Abstract

- Diabetic Retinopathy (DR) is an ophthalmic condition associated with diabetes mellitus which is caused by high blood sugar levels damaging the back of the eye (retina) via progressions inside the veins of the retina. Currently Diabetic Retinopathy is still screened as a three-stage classification which is a tedious strategy and along these lines this paper focuses on developing an improved methodology. In the present methodology, we taught a convolutional neural network form on a major dataset including around 45 depictions to mathematically analyse and characterize extreme goals i.e. hyper reflective foci (HRF) of the retina dependent on their seriousness. In this paper, DR is constructed which takes the enter parameters as HRF fundus photo of the eye. Our experimental data set is generated by processing the data provided by 301 hospitals. The experimental results show that random forest (RF) in the machine learning model can attain 80 Percentage of accuracy. Three classes of patients are considered - healthy patients, diabetics retinopathy patients and glaucoma patients. An informed convolutional neural system without a fully connected model will also separate the highlights of the fundus pixel with the help of the enactment abilities like relu and softmax and arrangement. The yield obtained from the convolutional neural network (CNN) model and the patient data achieves an institutionalized 97 percentage accuracy. The resulting methodology is therefore of great potential benefiting to ophthalmic specialists in clinical medicine in terms of diagnosing earlier the symptoms of DR and mitigating its effects.

Item Type: Article
Schools: Schools > School of Computing, Science and Engineering
Journal or Publication Title: Journal of Mechanics in Medicine and Biology
Publisher: World Scientific Publishing
ISSN: 0219-5194
Depositing User: OA Beg
Date Deposited: 11 Jun 2021 09:44
Last Modified: 11 Jun 2021 09:45
URI: http://usir.salford.ac.uk/id/eprint/60926

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)

Downloads

Downloads per month over past year