The study showed that artificial intelligence can be used to provide widespread, cost-effective eye screenings via telemedicine to assist ophthalmologists in improving vision outcomes.
Grand Warszawski / Shutterstock.com
A new study conducted by researchers from Genentech and Roche shows first-time proof that artificial intelligence can detect the severity of diabetic macular edema, which is a leading cause of blindness.
On Monday, researchers from Genentech and its parent company Roche published the study, Deep Learning Predicts OCT Measures of Diabetic Macular Thickening From Color Fundus Photographs” in the journal, Investigative Ophthalmology & Visual Science. The study showed that artificial intelligence can be used to provide widespread, cost-effective eye screenings via telemedicine to assist ophthalmologists in improving vision outcomes for millions of people with diabetes who may not be getting regular eye exams. The article is the first to be published that is part of a Roche/Genentech’s Ophthalmology Personalized Healthcare initiative. The initiative, Roche said in a statement, aims to combine meaningful large-scale data and AI technology to predict and prevent ocular conditions and preserve vision.
The study, Roche said, adds to the growing amount of scholarly information about the use of artificial intelligence in ophthalmology. The study sheds light on how a company such as Roche and its Genentech subsidiary can use its clinical trial database to develop AI algorithms to predict the presence of disease, risk of disease progression, and response to treatment; all of which could be supplied to ophthalmologists to deliver higher quality personalized healthcare.
Diabetic macular edema (DME) is a condition in which the retina develops diabetic microangiopathy with subsequent accumulation of fluid in the macula. It is a leading cause of vision impairment in people with diabetes, compromising their function and quality of life. According to the article, in 2017, approximately 425 million people worldwide had diabetes, and this number is estimated to grow to 629 million by 2045. Adults with diabetes and DME also have a substantially higher risk of cardiovascular morbidity, mortality, and amputation risk than those without, creating a further public health hazard.
According to the American Academy of Ophthalmology, having a dilated eye exam yearly or as recommended by an ophthalmologist can prevent 95 percent of diabetes-related vision loss. Roche though, said that many diabetic patients do not receive these kinds of examinations annually. Because of that, diagnoses of DME, or the severity of the disease, go unknown in many individuals. Because of that, many people with DME may be under-diagnosed or treated late and are at risk for irreversible vision loss or blindness, Roche said.
Roche and Genentech said that the study was the first time that researchers were able to demonstrate that artificial intelligence, specifically deep learning technology, is able to detect swelling in the macula, which is the part of the eye responsible for central vision and allows for people to see color and details. Also, the study showed that DLT can detect the severity of the swelling in people with diabetes. This swelling in the macula is known as diabetic macular edema (DME), a sight-stealing condition that impacts approximately 10 percent of the 425 million people around the world living with diabetes,” Roche said.
The primary way of diagnosing diabetic macular edema is through the use of optical coherence tomography (OCT), which takes three-dimensional, cross-sectional images of the macula. However, in its announcement, Roche said OCT is not often used in screening programs or telemedicine. OCT uses two-dimensional color photos called color fundus photos. Roche said one of the issues of these types of images is the two-dimensional nature of CFPs actually makes detecting the severity of DME difficult. In order to address this limitation, the Roche/Genentech research assessed how deep learning can automatically view CFPs to accurately detect DME and determine its severity.
Roche said its researchers used approximately 18,000 CFPs and associated OCT images during Genentech’s previous Phase III DME studies to develop and assess the performance of deep learning algorithms. The researchers said the results of the study showed that “the best deep learning algorithm was up to 97 percent accurate in detecting DME severity using CFP images alone.”
“Such results underscore the promising potential of AI in increasing screening capacity via telemedicine with appropriate triage to assist ophthalmologists in improving vision outcomes for a large population of patients who may not be getting comprehensive eye exams,” Roche said in its announcement.
