Artificial intelligence (AI) has been making waves for some years; those waves are continuing to affect the world of medicine. The FDA has recently fast-tracked the approval of a diagnostic system known as IDr-DR. This system can detect greater than mild levels of diabetic retinopathy, a leading cause of blindness among diabetics.
Diabetic retinopathy (DR) is a common disease among the diabetic community and is the leading cause of vision loss in these patients. Increased blood sugar levels damage the blood vessels in the eyes by causing them to become damaged and leak. In advanced cases, called proliferative diabetic retinopathy, the body creates new blood vessels which are not as strong as the originals and easily break, allowing more fluid into the retina. This release of fluids can lead to retinal scar tissue and potentially retinal detachment.
For patients with diabetes, it is extremely important to have annual eye exams for vision loss prevention and to keep your treatment options open.
This revolutionary machine is an artificial intelligence-based diagnostic system, the first of its kind to gain FDA approval. Developed by the world’s leading retina specialists, this advancement in diagnostics could help identify diabetic retinopathy before it does permanent damage.
Your optometrist uses a fundus camera to capture 2 detailed images of each eye. These cameras have a built-in microscope that allows them to capture your retina, optic disc, and macula. These images are then uploaded into the IDx-DR program on a computer. This program goes to work, analyzing the images for signs of retinopathy. In clinical studies of over 900 images, the IDx-DR correctly detected diabetic retinopathy in 87% of cases and correctly detected the absence of DR in 90% of those instances.
A microscopic image of the eye, as taken by the fundus camera (right).
Typically, diabetic retinopathy diagnosis requires an optometrist or ophthalmologist to screen for early diagnosis. The IDx-DR gives practitioners its own early input on the eye based on its algorithm. Results are not definitive; the program merely gives one of two answers:
These two responses inform the patient whether they should be seeking additional attention based on their results for this condition. In this way, there is the increased potential for patients with diabetes to monitor their ocular health without having to seek a specialist in DR. If they do begin to show signs of this disease, they can quickly seek treatment, mitigating long-term (and permanent) side-effects.
While this new technology may create ease of diagnosis for the bulk of those with diabetes, it does not take the place of an optometrist, nor is everyone a good candidate for the IDx-DR. You should not be screened by the device if you have any conditions including blurred vision, floaters, previously diagnosed macular edema, and more. Those who have undergone laser treatment, eye surgery, or any eye injections should also not use the device.
The advent of machines or programs like the IDx-DR could pave the way for more technology like this in-patient care. AI can alleviate some healthcare system strain by potentially lowering costs, improving productivity, and limiting barriers to treatment. This type of testing is not yet widely available in Canada, so please remember the importance of routine exams with your optometrist to determine any diabetic retinal changes.
If you’d like to know more about diabetic eye disease and its prevention, read How the Optomap Retinal Scanner Helps Your Eyes or Understanding the Basics of Retinal Vein Occlusion.