AI Detects Diabetic Retinopathy in Real-Timedec 24th zodiac sign

AI Detects Diabetic Retinopathy in Real-Timedec 24th zodiac signKey points Diabetic retinopathy is a leading cause of preventable blindness. One in four American adults with vision loss experiences anxiety or depression, according to the CDC. Research shows that a deep-learning system can now deliver real-time diabetic retinopathy detection. Source: TobiasD/Pixabay

By 2050, the National Institute of Health (NIH) National Eye Institute estimates that 14.6 million Americans will have diabetic retinopathy. A new study published in The Lancet demonstrates how artificial intelligence (AI) machine learning can screen in real-time for diabetic retinopathy—a leading cause of preventable blindness, particularly in areas with low-income or middle-income economies.

According to the Centers for Disease Control (CDC), one in four American adults with vision loss reported anxiety or depression. Moreover, vision loss has been linked to fear, anxiety, worry, social isolation, and loneliness. Scientists affiliated with Google Health and their collaborators applied artificial intelligence (AI) machine learning to detect one of the most common causes of preventable blindness—diabetic retinopathy.

Diabetic retinopathy is a common microvascular complication of diabetes. Diabetes is a group of metabolic diseases that negatively impacts how your body converts food into energy, resulting in persistently elevated levels of sugar (glucose) in the blood (hyperglycemia). Specifically, a person with diabetes is either unable to effectively use the insulin produced by the body or the pancreas fails to produce adequate insulin. Diabetes is a major cause of blindness, stroke, lower limb amputations, heart attacks, and kidney failure, according to the WHO.

Globally there are an estimated 537 million adults living with diabetes—a figure that is projected to increase to 783 million by 2045, according to the International Diabetes Federation (IDF). If diabetes is not controlled, it can lead to a number of serious health issues. In 2019, according to figures from the World Health Organization (WHO), 1.5 million people worldwide died due to diabetes.

“To our knowledge, this prospective study is the largest real-world validation to-date of a deep-learning system for eye screenings, having screened more than 7000 patients across both urban and rural care settings,” wrote the researchers with affiliations with Google Health, Verily Life Sciences, Imperial College London, Sankara Nethralaya, Nakornping Hospital, Vajira Hospital, and Rajavithi Hospital.

“A deep-learning system can deliver real-time diabetic retinopathy detection capability similar to retina specialists in community-based screening settings,” the researchers wrote.

There were over 7,600 eligible study participants and over 2,400 patients were referred for low visual acuity, diabetic macular edema, ungradable images, or diabetic retinopathy. The deep learning algorithm had an accuracy of 94.7 percent and a confidence interval of 95 percent for vision-threatening diabetic retinopathy.

“The deep-learning system enabled instantaneous diabetic retinopathy or diabetic macular edema grading assessments and referral recommendations to patients with diabetes who typically would have had to wait weeks before receiving their screening results due to the scarce availability of eye doctors,” the researchers reported.