Could an app that analyzes cellphone pictures of the eye replace the diabetes finger prick?
(ED NOTE: The next Killer App of Digital Healthcare may be the non-invasive glucose monitor, and who knows if optics, nanotech, or the following app will do the trick?)
October 17, 2013
Bloodless glucose monitoring: many have tried, none have succeeded. At least yet.
There’s no question why the diabetes community would want to ditch the daily lancet and test-strip routine for a cheaper, more convenient method of monitoring blood sugar levels.
In his lab at University of Toledo, bioengineering professor Brent Cameron has spent years studying ways to use light to non-invasively measure blood glucose. Medical device company Freedom Meditech, which developed an ophthalmic medical device that would enable eye care professionals to screen patients for diabetes during routine exams, licensed some of his technology several years ago.
Now a Toledo-based startup called IRISense LLC is in the early stages of commercializing more of his research in the form of a smartphone app that would predict blood glucose levels based on images of the eye.
But why the eye, rather than the fingertip or theearlobe? There’s glucose in the aqueous humor, a transparent, gelatinous fluid between the cornea and the iris, and there’s also no skin in the way, explained CEO Brent Cousino. There is, however, a small delay between glucose levels in the blood and glucose levels in the aqueous humor that seems to be about five minutes.
But that doesn’t seem to be causing too much trouble. Cameron developed an algorithm to predict glucose levels in the aqueous humor based on changes in the shape of the iris captured through by a cellphone camera. The idea behind IRISense is that it will develop an app that users would first have to calibrate by manually entering blood glucose levels. Then, they would take a picture of their eye whenever they need a blood sugar reading. Their readings would be displayed and automatically stored over time.
The company, which was spun out of University of Toledo and is owned by UT Innovation Enterprises, has developed a beta app that captures the images needed to apply the algorithms, and has tested it in a cohort of 17 people, Cousino said. Next, it needs to automate the image capture process and further refine the algorithm. That will involve a second study — this one led by a primary investigator who’s an endocrinologist, he said.
Ideally, a strategic partner or potentially some angel investors would follow that study. Since the eventual goal is to replace current testing methods, the app would also need FDA approval.
Using the eye to measure glucose isn’t a new idea. Fovi Optics and Oculir are among the companies that have tried, and failed, to commercialize eye-scan devices for glucose monitoring, either because they were not accurate or not practical. Cousino certainly isn’t na ve to that, or to the fact that there are a number of other companies taking different approaches to non-invasive monitoring, too.
“We’ve looked at the competing products and we know what’s out there,” he said. “We still feel good about what we’ve got.”