Sankara Eye deploys AI-based tool to diagnose diabetic retinopathy

Our Bureau Ahmedabad | Updated on June 01, 2021

Cloud-based Netra.AI uses deep learning to identify retinal conditions


The latest convergence of artificial intelligence (AI) and retinal diagnosis has given a new hope to those at risk of diabetic retinopathy. Sankara Eye Foundation and Singapore-based Leben Care have deployed a comprehensive retina risk assessment software - Netra.AI to enable faster and accurate detection of retinal disorders in large populations with limited healthcare resources.

This cloud-based AI solution is powered by Intel technology and uses deep learning to identify retinal conditions in a short span of time with accuracy matching with human doctors. Diabetic Retinopathy is the major cause for blindness and vision loss in working population.

With a large portion of diabetic population residing in rural India, there is a concern on shortage of trained ophthalmologists to diagnose diabetic retinopathy.

Early diagnosis

Kaushik Murali, President Medical Administration, Quality & Education, Sankara Eye Foundation informed that there is a gap between the onset of disease and it reaching a stage, where it becomes irreversible blindness. “If (by using AI), we are able to diagnose diabetic retinopathy early, we can intervene at that stage and try to either modify the disease so that the retinopathy itself doesn’t progress, or can preserve the limited vision for the patient,” he said.

“We tried to mimic what we were doing as a clinician and be able to look at AI algorithm. That enabled us to get a very high sensitivity (99.7 per cent) and specificity (98.5 per cent) - similar to how ophthalmologist would diagnose diabetic retinopathy,” said Murali. The report is generated within a few minutes thereby helping doctors to provide instant counsel for patients needing to be referred to the hospital.

Netra.AI analyses images from portable, technician-operated fundus camera devices. This helps to get referable DR grading via a cloud-based web portal immediately. The solution uses AI algorithms, developed in collaboration with leading retina experts, with a four-step deep convolutional neural network (DCNN). Netra.AI was validated at Sankara Eye Hospital with internal and external validation images from the patients. The neural network helps detect the stage of the disease and annotate lesions based on pixel density in the fundus images.

Netra.AI can also diagnose other retinal conditions like glaucoma, macular degeneration and other retinal pathologies.This reduces the screening burden on healthcare specialists.

Prakash Mallya, Vice-President & Managing Director – Sales, Marketing and Communications Group, Intel India, commented that the development was a “fantastic complement to ophthalmologist or healthcare specialist who are doing the diagnosis.”

Commenting on the emerging opportunities with AI, Mallya underlined the limitless possibilities ahead. “Due to Covid-19, the remote diagnostics has gone up. We see this trend dramatically continuing, and AI can make a significant impact. That, even beyond healthcare and in areas such aslike agriculture, education, smart cities and many others.”

Globally, Netra.AI has screened 3,293 patients worldwide so far and identified around 812 at-risk patients.

Published on May 31, 2021

Follow us on Telegram, Facebook, Twitter, Instagram, YouTube and Linkedin. You can also download our Android App or IOS App.

  1. Comments will be moderated by The Hindu Business Line editorial team.
  2. Comments that are abusive, personal, incendiary or irrelevant cannot be published.
  3. Please write complete sentences. Do not type comments in all capital letters, or in all lower case letters, or using abbreviated text. (example: u cannot substitute for you, d is not 'the', n is not 'and').
  4. We may remove hyperlinks within comments.
  5. Please use a genuine email ID and provide your name, to avoid rejection.