Using a Smartphone Attachment to Make Neurological Screenings Equitable

December 14, 2023 - Screening people early and often for potential health conditions is critical for stemming disease progression and improving patient outcomes. With the advent of advanced digital health tools, various types of screenings can be brought outside the hospital walls and into the community.

With this goal in mind, University of California San Diego (UC San Diego) engineers recently developed a smartphone attachment that can help screen people for a variety of neurological conditions. The attachment enables pupillometry, that is, measuring pupil dilation, which can support neurological assessments.

However, pupillometers are expensive, creating a major barrier to neurological screenings outside of certain healthcare facilities, said Edward Wang, an electrical and computer engineering professor at UC San Diego, in an interview with mHealthIntelligence.

“Some clinics have it, most clinics don't,” he said. “More importantly, you won't have this at home. This is not a blood pressure monitor where you can at least go try to buy one. You're not going to buy one of these… And so our vision was, ‘Hey, why can't this be more available?’ And our hope was to get these devices, and the smartphone, to provide this measurement.”

Using smartphones to perform pupillometry is not novel, but health equity issues can arise when using smartphone cameras.

“People in our field have thought of the idea of using smartphones to do pupil measurements,” said Wang, who is also director of the Digital Health Technologies Lab in The Design Lab at UC San Diego. “The issue is that regular cameras cannot distinguish very well between dark irises and the pupil. A dark iris is much like how the skin is darker for some people because of the melanin.”

He further explained that the pupil acts like a ‘black hole,’ allowing light to pass through to the back of the eye, where most of the light is absorbed.

Clinical pupillometers use near-infrared light to measure the pupil, according to a paper published in Scientific Reports about the development and use of the smartphone attachment. The melanin of the iris reflects longer wavelengths and appears bright, contrasting with the dark pupil. However, near-infrared imaging is not possible with smartphone cameras because they have built-in filters that cut off light outside the visible spectrum, like infrared light, to improve image quality.

“People have tried just using the straight [smartphone] camera…they found is that, yeah, you can kind of [perform pupillometry], but you can only do it for blue eyes. [With] brown eyes, black eyes, it won't work,” Wang said.

Thus, Wang and his team set out to create a device that could enable accurate pupillometry via the smartphone camera regardless of skin color. The attachment fits over a smartphone’s camera, enabling a novel pupillometry technique that leverages the ‘far-red spectrum.’

“Light is a continuum…[and] there's a spectrum of red,” Wang explained. “Darker red is longer in wavelength; the longer the wavelength, the closer we get to that edge where you can't see. What we found is the range before you cross into this part where your eyes can no longer see. We called it far-red, so far into the spectrum of red, this range where the camera can still see it, but your eyes don't really see it very well.”

The engineers found that they could use this tiny segment of light in the far-red spectrum to shine onto the eye such that the camera can still see the light, but the melanin doesn't absorb it.

“The attachment that goes over the camera essentially blocks out the light outside of the far-red region into blue,” Wang said. “So, what we call a long-pass filter.”

Thus, using far-red light while blocking out other wavelengths makes the iris appear lighter, which allows the pupil to be seen more easily with a regular camera.

The development began in earnest about a year ago, and the team published the paper describing the attachment and testing its use in August 2023.

The study shows that the proposed pupillometry technique uses a convolutional neural network and measures pupil diameter as it changes in a frame-by-frame video. They compared the pupil-iris contrast of 12 participants across four smartphone models, finding that the contrast increases by an average of 451 percent with the novel technique. A validation study of 11 participants revealed that the prototype attachment achieved a mean absolute error of 2.4 percent during a pupillary light response test.

The development team’s next steps include making the prototype scalable by reducing the number of components to lower the cost of mass production. They are also considering using the device in a large study examining Alzheimer's Disease screenings.

“Our main investigation is [focused on identifying] pupillometry markers to provide insight into cognitive welfare,” Wang said.

Prior research has already found a correlation between pupil responses and various neurological diseases, including traumatic brain injury, Alzheimer’s Disease, Parkinson’s Disease, and mild cognitive impairments. The smartphone attachment could potentially bring screening for these diseases outside the hospital. However, the device must be user-friendly, especially for older populations, to be helpful in this space,

“Making whatever [the tool] is accessible to older people is necessary,” said Alison Moore, MD, chief of the geriatrics division in the department of medicine at UC San Diego Health, in an interview with mHealthIntelligence.  

Not only do device makers need to take into account older adults’ limitations, like the fact that they may have sensory impairments, but they must also consider the caregivers of older adults.

“It's important for people to help manage their own health or have their families help manage their health,” Moore said. “Thinking about not only the context of the individual but who their caregiver team is and thinking about devices that could help them help the older adult with their care [is important].”

And it’s not just elder care. The attachment also has potential utility in lower resource settings and situations.

For example, these attachments could be distributed among high school football coaches, who could use them to conduct concussion screenings on kids knocked down on the field, making equitable concussion testing easily accessible, Wang said.

As the engineering team continues to build the attachment and make it scalable, they aim to ensure that human-centered design with a focus on health equity remains at the center.

“We can talk about usability, we can talk about other things, but I think those are sort of not inherently the innovation,” Wang said. “The innovation is to really be inclusive.”