Michael Raizman, MD, of Ophthalmic Consultants of Boston, spoke with Ophthalmology 360 at the 2026 ASCRS Annual Meeting about an investigational screening device for keratoconus. Initial studies have shown positive outcomes, and this screening tool may enable clinicians to diagnose keratoconus earlier, before there is additional vision loss.
Michael Raizman, MD:
I’m Michael Raizman. I’m an ophthalmologist in Boston. I work at Ophthalmic Consultants of Boston and Tufts University School of Medicine. At this year’s annual ASCRS meeting, I presented a feasibility study on a new detection device to look at keratoconus. We know that keratoconus is significantly underdiagnosed. There’s lots of reasons for this, but one major reason is the fact that there’s no good screening devices that are widely accessible and inexpensive. Many doctors’ offices do not have topography or tomography available. It would also be nice if we could have screening outside of eye clinics, pediatric offices, schools, allergy offices, and around the world where there’s not ready access to medical care. These are all places where it would be great to have this type of detection device.
I presented a feasibility study looking at this device, along with my co-investigator, Dan Tran in Irvine, California. This initial feasibility study is being followed by 2 additional studies. The device projects Meijer images onto the cornea and then creates a surface map based on those images. The device will, when it’s fully developed, have an algorithm that will tell us whether there is a high likelihood of keratoconus or not. The device is not intended to be a diagnostic device, but rather a screening device to help us find patients who need further diagnostic testing.
In this initial study, we enrolled healthy eyes as well as patients with keratoconus. We stratified the keratoconus into mild, moderate, and severe. We defined keratoconus by traditional methods, so we were sure that we were looking at classic keratoconus cases in this study.
We were really pleased to see that we were able to capture images quite readily; 139 out of 140 eyes had good capture of images. The single eye that couldn’t be captured had severe distortion of the ocular surface and obvious keratoconus. This would not be an eye we would normally need to screen; it would be one that would be an obvious case of keratoconus. We were also very pleased with the ease of use of the device. We had the user rate how easy it was or how difficult it was to capture the images on a scale of 1 being very difficult or the user being dissatisfied with the use of 5 being very satisfied with the ease of use of the device, and the mean score was about 4.
The eyes that had severe keratoconus took a little bit longer to capture, but despite that, the rate of capture is very quick as well, under 5 seconds for the capture of an image across all of our study eyes. We are currently conducting additional studies to follow-up on this feasibility study. The next study, which is nearly completed, will look at the images generated and how we can make an algorithm that will allow us to decide if a patient should have subsequent testing for keratoconus.
We’re also using machine learning so that the device can become better at detecting keratoconus suspects. If we’re successful, this device could be a very useful tool in providing better screening in multiple settings. It’s very ergonomic, easy to use. Now that we have an epi-on procedure for treating keratoconus, it’s especially important that we make the diagnosis early. We need to have better screening, find patients at an earlier stage of the disease. We’d rather treat the disease when the vision is still good, rather than, as we do now, often finding the patients when they’ve already lost some vision. This device should allow us to accomplish that.
Related Content
