How AI is Changing Diabetic Retinopathy Diagnosis

by | May 29, 2018 | Health Featured

Technology has shifted and advanced the way things are done in many different areas, with the most well-known in business operations. Not two decades ago, consumers had to venture out to their favorite retail store to buy the products they wanted to or needed, but today, the idea of at-home, online shopping is the norm. Even though technology has infiltrated the business arena is vast ways, innovation is beginning to touch healthcare in a similar fashion. Although there are many implications for digital enhancements throughout the healthcare arena, artificial intelligence has shown the most significant promise in medicine over the last few years.  That charge had been led by none other than the technology giant, Google.

A recent article highlighted the movement toward technological advances in medicine, focused on Google’s research and artificial intelligence arm, DeepMind. The firm set its target on reducing a well-known condition that impacts a large number of adults with diabetes each year throughout the UK and beyond. Diabetic retinopathy, a health issue of the eye that can lead to distorted vision and eventual blindness, must be caught early and treated quickly in an effort to maintain healthy eyesight. DeepMind’s project in this area has shown great promise in improving diabetic patients’ ability to avoid a late diagnosis or missed diagnosis of retinopathy as their underlying condition progresses. Similar projects are currently being greenlighted through the FDA in the United States, with artificial intelligence at the forefront.

How Diabetic Retinopathy is Diagnosed

Diabetic retinopathy takes place when blood vessels of the retina are damaged over time. In the early stages of retinopathy, individuals with diabetes may not experience any major symptoms, but as the condition worsens, there may be mild to severe vision problems that occur. Some people with diabetes have blurred or fluctuating vision, impaired color vision, dark or empty areas, and eventual blindness when the retinopathy goes untreated. Both Type 1 and Type 2 diabetics may ultimately be diagnosed with diabetic retinopathy, and the risk is higher when blood sugar is poorly managed.

Diagnosing diabetic retinopathy requires a visit to the doctor when vision becomes spotty or hazy, or when there are sudden, drastic changes in eyesight. In most cases, these symptoms trigger a full evaluation of the retina by a doctor or a specialist to determine the cause. Diagnosing diabetic retinopathy requires an exam focused on the health of the retina, specifically any signs of abnormal blood vessels, swelling or fatty deposits, scar tissue, or retinal detachment. The exam may also include other tests while the eyes are dilated. However, individuals with diabetes are best protected from blindness due to diabetic retinopathy when advanced screenings take place before eyesight becomes a problem.

Improvements with Artificial Intelligence                                         

Getting screened for eye disorders as a person with diabetes is crucial to maintaining a high level of health, no matter how well blood sugar is managed over time. A team of medical negligence solicitors explains that when diabetic retinopathy goes overlooked or undiagnosed for an extended period, additional complications may arise that lead to blindness. Some doctors do not automatically screen for the disorder during routine wellness checks, due in part to the time crunch they experience in seeing many patients throughout the day. If an individual with diabetes does not know the warning signs of retinopathy and does not bring up concerns with his or her GP, a diagnosis may go unnoticed and ultimately lead to a total loss of eyesight. While retinopathy is not completely avoidable, having a more efficient and accurate method to perform screenings is beneficial to people with diabetes.

Artificial intelligence through projects like Google’s DeepMind is promising because of the time reduction it offers to busy healthcare providers throughout public and private facilities. In trial results, the retinopathy screening device used by DeepMind has shown the potential to detect early signs of eye disease faster than a human specialist and with a great deal of accuracy. The AI screening is able to champion this task through analysing medical images and retinal scans so that the underlying algorithm used to provide a diagnosis is as smart as possible. The technology still needs to go through formal clinical trials before it can be used in the open healthcare market, but the early results offer hope for individuals with diabetes.

Share This