For the first time, the US Food and Drug Administration has approved an artificial intelligence diagnostic device that doesn’t need a specialized doctor to interpret the results. The software program, called IDx-DR, can detect a form of eye disease by looking at photos of the retina.
It works like this: A nurse or doctor uploads photos of the patient’s retina taken with a special retinal camera. The IDx-DR software algorithm first indicates whether the image uploaded is high-quality enough to get a result. Then, it analyzes the images to determine whether the patient does or does not have diabetic retinopathy, a form of eye disease where too much blood sugar damages the blood vessels in the back of the eye. Diabetic retinopathy is the most common vision complication for people with diabetes, but is still fairly rare — there are about 200,00 cases per year.
In one clinical trial that used more than 900 images, IDx-DR correctly detected retinopathy about 87 percent of the time, and could correctly identify those who didn’t have the disease about 90 percent of the time.
The software is unique because it’s autonomous and there’s “not a specialist looking over the shoulder of [this] algorithm,” IDx-DR founder Michael Abràmoff told Science News. “It makes the clinical decision on its own.” This means that the technology can be used by a nurse or doctor who’s not an eye specialist, making diagnosis more accessible. For example, patients wouldn’t need to wait for an eye specialist to be available to get a diagnosis.
IDx-DR is part of a growing trend of algorithms learning how to spot and diagnose disease. Earlier this year, scientists trained a different algorithm to learn how to recognize conditions including age-related vision loss and diabetic retinopathy. Google, too, is training its DeepMind AI to spot eye disease. Now that the FDA has cleared IDx-DR, it might lead the way to a new slew of autonomous diagnostic tests and the trade-offs they bring. These diagnoses could be more convenient for patients (and possibly even more accurate than doctors). But of course, not having a specialist “looking over the shoulder,” as Abràmoff puts it, raises the question of who will be responsible when the diagnosis is wrong.
The official FDA website states that “The FDA evaluated data from a clinical study of retinal images obtained from 900 patients with diabetes at 10 primary care sites. The study was designed to evaluate how often IDx-DR could accurately detect patients with more than mild diabetic retinopathy.”
In the study, IDx-DR was able to correctly identify the presence of more than mild diabetic retinopathy 87.4 percent of the time and was able to correctly identify those patients who did not have more than mild diabetic retinopathy 89.5 percent of the time.
Patients who have a history of laser treatment, surgery or injections in the eye or who have any of the following conditions should not be screened for diabetic retinopathy with IDx-DR: persistent vision loss, blurred vision, floaters, previously diagnosed macular edema, severe non-proliferative retinopathy, proliferative retinopathy, radiation retinopathy or retinal vein occlusion. IDx-DR should not be used in patients with diabetes who are pregnant; diabetic retinopathy can progress very rapidly during pregnancy and IDx-DR is not intended to evaluate rapidly progressive diabetic retinopathy. IDx-DR is only designed to detect diabetic retinopathy, including macular edema; it should not be used to detect any other disease or condition. Patients will still need to get a complete eye examination at the age of 40 and at the age of 60 and also if they have any vision symptoms (for example, persistent vision loss, blurred vision or floaters).
IDx-DR was reviewed under the FDA’s De Novo premarket review pathway, a regulatory pathway for some low- to moderate-risk devices that are novel and for which there is no prior legally marketed device. IDx-DR was granted Breakthrough Device designation, meaning the FDA provided intensive interaction and guidance to the company on efficient device development, to expedite evidence generation and the agency’s review of the device. To qualify for such designation, a device must provide for more effective treatment or diagnosis of a life-threatening or irreversibly debilitating disease or condition, and meet one of the following criteria: the device must represent a breakthrough technology; there must be no approved or cleared alternatives; the device must offer significant advantages over existing approved or cleared alternatives; or the availability of the device is in the best interest of patients.
The FDA is permitting marketing of IDx-DR to IDx LLC.
The FDA, an agency within the U.S. Department of Health and Human Services, protects the public health by assuring the safety, effectiveness, and security of human and veterinary drugs, vaccines and other biological products for human use, and medical devices. The agency also is responsible for the safety and security of our nation’s food supply, cosmetics, dietary supplements, products that give off electronic radiation, and for regulating tobacco products.
What it does: The software is designed to detect greater than a mild level of diabetic retinopathy, which causes vision loss and affects 30 million people in the US. It occurs when high blood sugar damages blood vessels in the retina.
How it works: The program uses an AI algorithm to analyze images of the adult eye taken with a special retinal camera. A doctor uploads the images to a cloud server, and the software then delivers a positive or negative result.
No specialist required: The FDA recently cleared AI-based software to help detect stroke, too. But the agency says this is the first device authorized to provide a screening decision without the need for a specialized doctor to interpret the image or results.
A look ahead: In a series of tweets, FDA commissioner Scott Gottlieb hinted that more AI devices could get the agency’s seal of approval soon. Gottlieb said the FDA is “taking steps to promote innovation and support the use of artificial intelligence-based medical devices.” Not to worry, though: AI diagnostics probably won’t be replacing doctors or other medical professionals anytime soon.