A research team at the University of Bonn, Germany (Rin Friedrich-Wilhelms Bonn University) has developed a technology that can investigate blindness and diabetic retinopathy (DR) symptoms caused by obstruction of retinal blood vessels due to diabetes with a smartphone.
In the United States, 40 to 45% of diabetic patients are said to have retinopathy. If you can quickly find the symptoms of retinopathy and take appropriate treatment, you can prevent blindness.
In cooperation with Sankara Eye Hospital in Bangalore, India, the research team has investigated a wide range of patient health conditions to find ways to quickly detect retinopathy early, even in developing countries that do not yet have a health diagnostic system. The research team focused on smartphones as devices that enable medical diagnosis at low cost. Smartphones are now widely distributed to developing countries to some extent and provide advanced camera functions and sensors. We tried to access some retinal imaging with the smartphone built-in camera and developed a special camera adapter to get clear images.
Maximilian Wintergerst, a professor of ophthalmology at the University Medical Center, said that the adapter was able to detect the lesion with 80% probability even in the early stages of retinopathy.
This diagnostic technique works by recording retinal changes. An ophthalmologist can do this in a minute or two with this technique. Above all, since it uses a smartphone, even a beginner can take a diagnosis image with a little training, and the image can be sent to a remote ophthalmologist through the Internet for diagnosis.
The research team is developing an application program to make this telemedicine easier. The application program encrypts and stores the image of the patient’s retina and the results of the doctor’s analysis, and is expected to be the first step to automatically diagnose retinopathy with AI.
In developing countries that do not have a medical system, Google has conducted tests with Thai medical institutions in an attempt to promote the diagnosis of retinopathy using AI. However, the diagnostic imaging technology, environmental issues, and facilities were ankles on this side, and the desired outcome was not achieved. On the other hand, this method using a smartphone can be an alternative to solving the problem of retinopathy diagnosis in that it does not require much photographer skills and does not require expensive equipment. Related information can be found here .