AI shows promise in screening diabetic retinopathy: study

Artificial Intelligence (AI) may prove to be extremely beneficial to detect diabetic retinopathy in patients at an early stage, a new study has shown. The study, carried out among patients coming to Mumbai’s civic-run dispensaries, revealed 100% reliability of AI for screening diabetic retinopathy.

Published in the Journal of American Medical Association Ophthalmology last week, the study included a screening of 231 patients with diabetes. The images of their eyes were taken on Remidio Non-Mydriatic Fundus on Phone, a portable device attached to a smartphone equipped with retinal imaging. The Medios AI on the device helped in detecting diabetic retinopathy and prompting health workers to refer patients for further care.

“There is huge potential in AI and this study highlights it,” said eye surgeon Dr. S. Natarajan, author of the study. He said conservative estimates show that nearly 12 million people with diabetes in India would have diabetic retinopathy at any point of time. “Innovative measures like these will decrease the burden of ophthalmologists in screening diabetic retinopathy,” he said.

The project was implemented by the Aditya Jyot Foundation for Twinkling Little Eyes (AJFTLE). Since the number of patients in the pilot study was small, AJFTLE decided to continue with the screenings. “Till July, we have screened 1,688 patients with diabetes, and the AI has been found to be sensitive in nearly 98% cases,” said Dr. Radhika Krishnan, chief executive officer of AJFTLE. The highlight of the AI is that it does not require an internet connection or electricity, gives the report in less than 30 seconds, and is entirely smartphone-based.

While the portable device was developed by Bengaluru-based Remidio Innovative Solutions, the Medios AI was developed by Medios Technologies, a subsidiary of Remidio based in Singapore. “The use of AI will enable screening for referable diabetic retinopathy in remote areas where services of an ophthalmologist are unavailable,” the study said, citing the need to study a large sample size to extend the results.

This article is closed for comments.
Please Email the Editor

Printable version | May 10, 2021 1:11:44 PM |

Next Story