New AI technology diagnoses eye diseases and assists the blind


Published: 2023-06-27 15:08

Last Updated: 2024-04-13 21:29

New AI technology diagnoses eye diseases and assists the blind
New AI technology diagnoses eye diseases and assists the blind

Novel artificial intelligence models are being used to diagnose eye conditions and aid blind people in their daily lives.

According to the World Health Organization (WHO), blindness affects more than 2.2 billion people worldwide, and in almost half of these cases, visual impairment could have been prevented or has yet to be addressed. Vision loss is only becoming more prevalent; projections show that it will increase by 55 percent in the next 30 years.

New technology has the ability to significantly alter the practice of eye medicine and improve the lives of millions of visually impaired people.

Identifying Eye Conditions

In the United Kingdom, age-related macular degeneration (AMD) is the most common cause of vision loss. Right now, there is no cure, and identifying AMD can be a lengthy process.

Konstantinos Balaskas, a consultant ophthalmologist at Moorfields Eye Hospital in London, said in an interview: “It takes a human expert 43 minutes to assess a dry AMD retinal scan.”

“This would put a huge strain on clinical resources.”

However, thanks to an AI-based “deep learning” system created by Balaskas and his colleagues at the National Health Service and University College London, medical professionals can identify dry AMD in two seconds, instead of 43 minutes.

The AI was trained on 5,000 retina scans classified by humans and validated on 900 scans from a different data set. “In a busy clinic, it would be impossible to do manually,” said Balaskas.

“This is the value of using AI.”

It is not just AMD that can be identified with the help of AI. In 2018, London researchers worked with AI company DeepMind, which is now owned by Google, to prove that AI can analyze eye scans and assess them for over 50 retinal diseases, performing as well as expert ophthalmologists.

Since then, researchers have worked to train and validate the system so that it works on all patients.

Pearse Keane, a researcher at University College London who is studying potential uses of AI, is confident that artificial intelligence will reduce the burden of chronic diseases like AMD on patients and lead to more efficient treatment.

Aiding the Visually Impaired

Be My Eyes is a mobile application that has aided over 250 million visually impaired people since 2012. By chatting with volunteers from around the world through the app, blind users can identify objects and better navigate the physical world.

With the release of GPT-4 earlier this year, a powerful large language model by OpenAI, Be My Eyes is testing a new iteration of the app that leverages artificial intelligence to chat instantly with users, answering questions and giving advice. With GPT-4, Be My Eyes analyzes users’ surroundings through their phone camera and provides guidance.

“Basic image recognition applications only tell you what’s in front of you,” Jesper Hvirring Henriksen, CTO of Be My Eyes, explained.

“They can’t have a discussion to understand if the noodles have the right kind of ingredients or if the object on the ground isn’t just a ball, but a tripping hazard—and communicate that.”