Signs of atrial fibrillation identified in the eye
UK researchers have analysed retinal imaging data from 90,000 people to detect signs of atrial fibrillation up to four years before clinical presentation
A new study published in PLOS Digital Health has outlined efforts to identify biomarkers for atrial fibrillation (AF) through artificial intelligence (AI) analysis of retinal images.
Research led by Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology analysed optical coherence tomography (OCT) scans and retinal photographs of 90,000 people from two large datasets (AlzEye and UK Biobank).
Speaking with OT about the research, lead author and Moorfields Eye Hospital consultant ophthalmologist, Josef Huemer, highlighted that there was ganglion layer and inner plexiform layer thinning in patients with established AF.
He added that these layer changes were also present in patients without an AF diagnosis when the images were captured.
“On average, they would be diagnosed with atrial fibrillation up to four years later,” Huemer said.
The ophthalmologist observed that AF is a disease that is “notoriously difficult” to diagnose.
“Up to 40% of patients with atrial fibrillation are not aware that they have the disease. It is often a silent disease. Very often patients only find out they have the condition when they have suffered a stroke or they are diagnosed with heart failure,” Huemer said.
He shared that the research could one day lead to a non-invasive test that identifies patients at risk of AF earlier – enabling them to undergo long term cardiac monitoring.
“This is one of the most promising fields in oculomics, where we hope that we can improve patient health outcomes and healthy lifespan,” he said.
Huemer noted that while there are wearable devices that are able to identify AF, these tools have relatively low sensitivity – leading to false positive referrals.
“That is something that we really want to avoid. We don’t want to create an additional burden on the system or increase anxiety in our patients,” he said.
Huemer highlighted that further tests are always needed to confirm a diagnosis of AF.
“In the context of atrial fibrillation, which is such an insidious disease and such a difficult disease to detect, if we can hasten the application of long-term cardiac monitoring, this could be really beneficial,” he said.
- Explore more topics
- Research
- Imaging
- Artificial intelligence
Comments (0)
You must be logged in to join the discussion. Log in