A team of scientists say that they have developed an artificial intelligence (AI) tool capable of diagnosing and predicting the risk of developing multiple health conditions — from ocular diseases to heart failure to Parkinson’s disease — all on the basis of people’s retinal images.
According to a peer reviewed study published by Nature, the world’s leading multidisciplinary science journal, the scientists have trained AI tools to detect disease using retinal images before, but what makes the new tool — called RETFound — special is that it was developed using a method known as self-supervised learning.
They explain that this means they did not have to analyse each of the 1.6 million retinal images used for training and label them as ‘normal’ or ‘not normal’, for instance.
“Such procedures are time-consuming and expensive, and are needed during the development of most standard machine-learning models,” they highlighted.
This is why instead, the scientists used a method similar to the one used to train large-language models such as ChatGPT.
“That AI tool harnesses myriad examples of human-generated text to learn how to predict the next word in a sentence from the context of the preceding words. In the same kind of way, RETFound uses a multitude of retinal photos to learn how to predict what missing portions of images should look like,” their study notes.
Pearse Keane, an ophthalmologist at Moorfields Eye Hospital NHS Foundation Trust in London who co-authored the study further described how the new tool works. “Over the course of millions of images, the model somehow learns what a retina looks like and what all the features of a retina are, this forms the cornerstone of the model, and classifies it as what some call a foundation model, which means that it can be adapted for many tasks,” he said.
The scientists also explained that a person’s retinas can offer a window into their health, because they are the only part of the human body through which the capillary network, made up of the smallest blood vessels, can be observed directly.
“If you have some systemic cardiovascular disease, like hypertension, which is affecting potentially every blood vessel in your body, we can directly visualize [that] in retinal images,” Keane said while adding that retinas are also an extension of the central nervous system, sharing similarities with the brain, which means that retinal images can be used to evaluate neural tissue. “The rub is that a lot of the time people don’t have the expertise to interpret these scans. This is where AI comes in,” the study co-author noted.
The team of scientists discloses that once they had pre-trained RETFound on those 1.6 million unlabelled retinal images, they could then introduce a small number of labelled images — say, 100 retinal images from people who had developed Parkinson’s and 100 from people who had not — to teach the model about specific conditions as the model had learnt from all the unlabelled images what a retina should look like according to the co-author and this means that it is able to easily learn the retinal features associated with a disease.
The researchers further observe that the system performed well at detecting ocular diseases such as diabetic retinopathy.
“On a scale where 0.5 represents a model that performs no better than a random prediction and 1 represents a perfect model that makes an accurate prediction each time, it scored between 0.822 and 0.943 for diabetic retinopathy, depending on the data set used. When predicting the risk for systemic diseases — such as heart attacks, heart failure, stroke and Parkinson’s — the overall performance was limited, but still superior to that of other AI models,” they explained.
They further point out that RETFound is so far one of the few successful applications of a foundation model to medical imaging.
So where do they go from here?
According to Nature, they are now looking ahead to what other types of medical imaging the techniques used to develop RETFound might be applied to.
Radiologist Curtis Langlotz, director of the Center for Artificial Intelligence in Medicine and Imaging at Stanford University, in California said that it will be interesting to see whether these methods generalize to more complex images for example to magnetic resonance images or computed tomography scans, which are often three- or even four-dimensional.
The authors of the study have made the model publicly available and hope that groups around the world will be able to adapt and train it to work for their own patient populations and medical settings.
“They could potentially take this algorithm and fine-tune it, using data from their own country to have something that’s more optimized for their use,” the co-author said.
