Google's AI program: Building better algorithms for detecting eye disease

March 13, 2018, American Academy of Ophthalmology
The ability of artificial intelligence (AI) to help screen patients for a common diabetic eye disease gains momentum with a new study published online today in Ophthalmology, the journal of the American Academy of Ophthalmology. Credit: Shutterstock

The ability of artificial intelligence (AI) to help screen patients for a common diabetic eye disease gains momentum with a new study published online today in Ophthalmology, the journal of the American Academy of Ophthalmology. Lily Peng, M.D., Ph.D., and her colleagues at Google AI research group, show that they could improve their disease detecting software by using a small subset of images adjudicated by ophthalmologists who specialize in retinal diseases. The specialists' input was then used to improve their computer's performance so that it was roughly equal to that of individual retinal specialists.

More than 29 million Americans have diabetes, and are at risk for , a potentially blinding eye . People often don't notice changes in their vision in the disease's early stages. But as it progresses, diabetic retinopathy usually causes vision loss that in many cases cannot be reversed. That's why it's so important that people with diabetes have yearly screenings.

In earlier research, Dr. Peng and her team used neural networks—complex mathematical systems for identifying patterns in data—to recognize diabetic retinopathy. They fed thousands of retinal scans into these neural networks to teach them to "see" tiny hemorrhages and other lesions that are early warning signs of retinopathy. Dr. Peng showed the software worked roughly as well as human experts.

But Dr. Peng is interested in developing a system that would be good enough for her grandmother. So, to improve the accuracy of the software, she included the input of retina specialists, ophthalmologists who specialize in diseases of the retina.

"For my grandma, I would love to have a panel of subspecialists who actually treat the disease, to sit and debate her case, giving their opinion," Dr. Peng said. "But that is really expensive and it's hard to do. So how do you build an algorithm that gets close to this?"

To tease out how this could be done, Dr. Peng compared the performance of the original algorithm with manual image grading by either a majority decision of three general ophthalmologists, or a consensus grading by three retinal specialists.

The grading of diabetic retinopathy can be a complex process that requires the identification and quantification of fine features such as small aneurysms and hemorrhages. As a result, there can be a fair amount of variability among physicians examining images, looking for disease.

The retina specialists graded the images separately, then worked together to resolve any disagreements. Their review and subsequent consensus diagnosis offered considerable insight into the grading process, helping to correct errors such as artifacts caused by dust spots, distinguishing between different types of hemorrhages, and creating more precise definitions for "gray areas" that make it difficult to make a definitive diagnosis. At the end of the process, the retina specialists indicated that the precision used in the decision process was above that typically used in everyday clinical practice.

Using these specialist-graded images, Dr. Peng could then fine-tune the software, which improved their model's performance and improved detection of disease.

"We believe this work provides a basis for further research and raises the bar for reference standards in the field of applying machine learning to medicine," Dr. Peng said.

Explore further: Study examines use of deep machine learning for detection of diabetic retinopathy

More information: Grader Variability and the Importance of Reference Standards for Evaluating Machine Learning Models for Diabetic Retinopathy, DOI: 10.1016/j.ophtha.2018.01.034 , http://www.aaojournal.org/article/S0161-6420(17)32698-2/fulltext

Related Stories

Study examines use of deep machine learning for detection of diabetic retinopathy

November 29, 2016
In an evaluation of retinal photographs from adults with diabetes, an algorithm based on deep machine learning had high sensitivity and specificity for detecting referable diabetic retinopathy, according to a study published ...

New laser scanners shed light on eye disease before vision loss occurs

December 22, 2017
SFU engineering science professor Marinko Sarunic has developed a high resolution retinal imaging scanner that will one day revolutionize eye care, helping ophthalmologists diagnose eye diseases before vision loss occurs.

Google AI can predict heart disease by looking at pictures of the retina

February 19, 2018
I can look into your eyes to see straight to your heart.

Researchers use deep learning to create an algorithm to detect a common diabetic eye disease

April 28, 2017
Researchers from the Byers Eye Institute at Stanford University have found a way to use artificial intelligence to fight a complication of diabetes that affects the eyes. This advance has the potential to reduce the worldwide ...

Annual eye exam is vital if you have diabetes

March 8, 2018
(HealthDay)—A yearly eye exam is a key part of diabetes treatment, experts say.

Recommended for you

Researchers are one step closer to developing eye drops to treat age-related macular degeneration

July 19, 2018
Scientists at the University of Birmingham are one step closer to developing an eye drop that could revolutionise treatment for age-related macular degeneration (AMD).

An orange a day keeps macular degeneration away: 15-year study

July 12, 2018
A new study has shown that people who regularly eat oranges are less likely to develop macular degeneration than people who do not eat oranges.

Injectable electronics offer powerful new tool in understanding how retinal cells work

June 28, 2018
Charles Lieber and his group are rewriting the rules of how scientists study retinal cells, and they're doing it with a single injection.

Why the eye could be the window to brain degeneration such as Alzheimer's disease

June 26, 2018
Researchers from Queen's University Belfast have shown for the first time that the eye could be a surrogate for brain degeneration like Alzheimer's disease (AD).

Microglia protect sensory cells needed for vision after retinal detachment

June 18, 2018
A research team at Massachusetts Eye and Ear has shown that microglia, the primary immune cells of the brain and retina, play a protective role in response to retinal detachment. Retinal detachment and subsequent degeneration ...

161 genetic factors for myopia identified

June 15, 2018
The international Consortium for Refractive Error and Myopia (CREAM) recently published the largest-ever genetic study of myopia in Nature Genetics. Researchers from the Gutenberg Health Study at the Medical Center of Johannes ...

0 comments

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.