How AI Will Transform Eyecare

I was in London recently to visit Moorfields Eye Hospital. Although it’s the oldest eye hospital in the world, there is nothing antique about it.

Moorfields is a world center for research and teaching, and is on the cutting edge of eyecare, from using stem cells to grow eyeballs to studying the human genome to develop personalized medicine.

While at Moorfields, I did a video interview with Dr. Pearse Keane, a research ophthalmologist who is heading a unique project in which Artificial Intelligence is being used to diagnose and predict the likelihood that a patient may have a retinal disease such as age-related macular degeneration or diabetic retinopathy.

Moorfields’ partner in this project is DeepMind Health, a division of DeepMind, which is a leading player in the AI field, and has developed algorithms that can read and analyze OCT (Optical Coherence Tomography) scans with speed and accuracy.

“OCT scans are well suited to this because it provides very high resolution images,” explained Dr. Pearse. “The axial resolution is about five thousandths of a millimeter, which is better than an MRI or CAT scan.”

Moorfields has assembled a huge database of scans that DeepMind’s algorithms are analyzing.

Using data from the study, researchers can see minute changes in the cellular structure of the eye enabling them to accurately monitor the progression of a disease such as AMD. Using data derived from the scans, research might be able to predict when AMD progresses from its dry form to its more serious wet form and offer early intervention.

Dr. Keane believes AI will transform both ophthalmology and eyecare. “I think the imagination is our limit in terms of where we can apply this,” he said.

You can see my interview with Dr. Keane at the VM Global Leadership Summit this week. It will then be available, along with other Summit content, on and in the April 24 issue.