Here is how it works:
A doctor uses a special retina camera to take a clear image of the retina of the patient. This image is sent to the IDx-DR software algorithm and it primarily checks if the image is clear enough(is high-quality or not) to get a good result. The software then analyzes the image to determine whether the patient has diabetic retinopathy or not. This is a form of eye disease that is prevalent in diabetic patients. Due to the high blood sugar levels, the blood vessels in the back of the eye get damaged. This disease is not that wide-spread with around 200,000 cases every year. In the recent clinical trial conducted using more than 900 sample images, the algorithm was able to successfully detect disease among 87 percent of the patients. It is able to successfully detect that 90 percent of the patients do not have this disease. MichaelAbràmoff, the founder of IDx-DR told Science News that ” The software is unique because it’s completely autonomous and there is not a single specialist looking over the shoulder of this algorithm”. He also said that ” It makes the clinical decisions completely on its own”. This enables the software to be used either by doctor or nurse or even anyone who doesn’t have a good expertise, making the diagnosis more accessible. With this, the patients o longer have to wait for a specialist to conduct the diagnosis. Source: The verge.
