Data Science Investment Counter —Funding raised by UK data science companies in 2018.
£ 5.640 Million

AI system from Google’s Deepmind is ‘as good as experts’ at detecting eye problems

A new artificial intelligence (AI) system can now identify 50 different types of eye conditions as accurately as a doctor, according to a new research published on Nature Medicine’s website.

Developed by Deepmind, one of Google’s AI companies, the system is the result of the research partnership with Moorfields Eye Hospital in London over the last 18 months. The collaboration aims to establish whether AI technology could aid clinicians when it comes to improving care for patients.

In the long term, we hope this will help doctors quickly prioritise patients who need urgent treatment – which could ultimately save sight.

– Mustafa Suleyman, DeepMind cofounder

The machine-learning program can analyse 3D retinal OCT scans for early signs of conditions like glaucoma, diabetic eye disease and macular degeneration. The researchers claim that, after 15,000 anonymous eye scans, the AI would be able to identify eye diseases with an accuracy rate of 94.5%.

“Our AI system can quickly interpret eye scans from routine clinical practice with unprecedented accuracy”, DeepMind cofounder Mustafa Suleyman said in a blog post on Monday, “It can correctly recommend how patients should be referred for treatment for over 50 sight-threatening eye diseases as accurately as world-leading expert doctors”.

DeepMind is planning to let NHS hospitals use a validated version of the technology for free for five years, under the conditions that the software meet all the necessary regulatory approvals. This is not the first time the London-based company invest in the NHS, with the original collaboration with Moorfields Eye Hospital’s sparking controversy, as over one million patient data were processed by DeepMind’s algorithm.

Moorfields said that, globally, over 285 million people live with some sort of sight loss. The potential of this kind of research could, therefore, be significant. According to the researchers, the next step would be to carry out clinical trials to assess how the technology could improve patient care in practice.

“These are early results,” said Mustafa, “but they show that our system could handle the wide variety of patients found in routine clinical practice. In the long term, we hope this will help doctors quickly prioritise patients who need urgent treatment – which could ultimately save sight.”

SHARE THIS ARTICLE:
BY SHACK15

Co-working space and blog dedicated to all things data science.

Subscribe to our newsletter