Where’s My Eye Scanner?

Over the past couple of decades, biometric sensors have slowly been integrated into our lives. The first fingerprint sensors were added to smartphones in 2007 and, since then, they have grown in popularity. Alongside facial recognition software, which was introduced into consumer devices in 2015, these technologies have firmly established the concept of people’s bodies securing devices.

However, when I look at the biometric devices currently in use and compare them with those featured in fiction, one device is conspicuous by its absence: the eye scanner. Once a sci-fi staple (think Blade Runner or Minority Report) the idea of our eyes being our primary identifier seems to have fallen out of favor. Will we ever see widespread use of eye scanners?

Retinal scanning is common, of course – made increasingly ubiquitous with OCT. This gold standard for non-invasive imaging of the retina is being constantly refined and repurposed to increase its range of functionality. One recent development in this area is spatio-temporal optical coherence tomography (STOC-T), which allows the identification of individual dead photoreceptors. But OCT devices tend to be relatively large and bulky – unlikely to fit into a compact modern smartphone any time soon.

Phones too continue to evolve – some researchers are developing a new application that turns a phone’s native technology into a pupillometer…

Does this mean that we’ll be scanning our eyes to get into our houses anytime soon? Probably not. Iris scanners have been incorporated into phones before, but the technology didn’t take off and was removed, as fingerprint and facial recognition systems are more reliable, are able to work in a wider variety of lighting and weather conditions, and – perhaps more importantly – carry no risk of eye damage. So, if the security application of eye scanning doesn’t seem likely in the immediate future, where might this technology be used instead?

Previously discussed retinal scans are regularly used diagnostically, but they could be taken to the next level with the use of artificial intelligence. Deep learning algorithms assessing fundus images have already been shown to make associations invisible to the human eye to predict factors about an individual, such as their biological sex, body composition, BMI, and waist-hip ratio (1). They’ve also been able to identify diseases, such as heart disease, Alzheimer’s disease, and childhood autism (2, 3, 4) at an early stage – potentially long before symptoms start to develop. As technology moves forward, it seems likely that such algorithms will further improve and find new and interesting associations. But no one likes a black box, and so researchers are working hard to address the issue of model explainability. Nevertheless, with AI and advanced retinal imaging, our eyes may increasingly act as a blueprint of our uniqueness…

Do you have any thoughts on or predictions for the future of retinal or iris scanning? If so, please send them via edit@theophthalmologist.com or comment below.