Application of Machine Learning, Neural Networks, and Deep Learning in Eye care Innovations

Artificial Intelligence methods have shown to be a promising tool in the field of medicine and more so in ophthalmology and Preventive eye care. AI is a field focused on automating intellectual tasks normally performed by humans, Machine Learning and Deep Learning are specific methods of achieving t...

Full description

Saved in:
Bibliographic Details
Published in:Indian journal of clinical biochemistry 2022-05, Vol.36 (S1), p.S32
Main Author: Ve, Ramesh S
Format: Article
Language:English
Subjects:
Algorithms
Artificial intelligence
Diabetic retinopathy
Innovations
Machine learning
Neural networks
Ophthalmology
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Artificial Intelligence methods have shown to be a promising tool in the field of medicine and more so in ophthalmology and Preventive eye care. AI is a field focused on automating intellectual tasks normally performed by humans, Machine Learning and Deep Learning are specific methods of achieving this goal. AI also includes rule based algorithms and convolutional neural networks (CNN) approaches for enhanced image analysis. Our diabetic retinopathy Image Screening research has incorporated DL image analysis algorithm and reports the validity of this automated retinopathy screening system. In clinical validation on over 1400 diabetic retinopathy patients the Algorithm showed a sensitivity of 90.75% (95CI 86.42, 94.62) and specificity of 97.75% (95CI 96.58, 98.60) respectively. Automated optic nerve head grading research incorporated CNN based concepts applied in staging glaucoma and shows accurate staging across fundus images of various quality. ML and DL methodologies are also applied in assessing visual field defects assessed using eye tracking and virtual reality based vision therapy projects. In this presentation we cover AI based approaches in our eye care innovations and better explain these concepts to a clinical audience with limited technical jargon.
ISSN:0970-1915
0974-0422