The U.S. Food and Drug Administration (FDA) recently authorized the use of a novel eye-tracking tool to assist in the diagnosis of autism in kids aged 16 to 30 months — a very early yet critical stage for discovering the condition in young children. While the new tech promises to help make identifying autism in toddlers speedier and more objective than current methods, helping bridge a national delay and gap in access to diagnosis, it cannot be a standalone tool.
Here’s everything you need to know about how and why eye tracking works to diagnose autism, and whether your child might benefit from it.
How Can Eye-Tracking Technology Diagnose Autism In Toddlers?
For the past 20 years, Ami Klin, Ph.D., director of the Marcus Autism Center of Children’s Healthcare of Atlanta, has been studying how autism can be detected in young children by analyzing where they focus their attention and what they look at.
“Looking behavior is a fundamental foundation for the way infants and toddlers acquire speech, language, and communication,” says Klin. “By observing others, they learn about their emotions, what they are referencing to in their behavior, and the emotional meaning of their actions.” He had a Ted Talk about this already in 2011.
Soon after, in 2013, Klin’s team found that kids as young as 2 months old who didn’t pay much attention to their caregiver’s eyes and mouth were later diagnosed with autism at the ages of 24 and 36 months. The steeper the decline in their interest in a caregiver’s facial expressions after the age of 2 months, and the more disturbed their “looking behavior,” the more severe their level of disability and diagnostic outcome was.
This is in line with a large, growing body of international research, and it’s these kinds of results that are now the basis of his team’s new biomarker-based technology: a small, portable, and wireless gadget named EarliPoint.
The device is a tablet that plays a 12-minute video portraying other toddlers as they interact — pointing at things and chatting away. While the young patient watches the video, the tool tracks their eye motions 120 times per second, monitoring where they’re looking and what social interactions most capture their attention. “The data collection procedure is as straightforward as watching TV,” says Klin.
An autistic child is less likely to pay attention to where the kids in the film are pointing, or what their faces are expressing.
How Reliable Is Eye Tracking For Autism?
When testing the tech on a group of 335 kids, some who were autistic and some who weren’t, the novel eye-tracking tool accurately pin-pointed 78% of the autistic kids as having autism and correctly identified about 85% of the non-autistic kids as not having autism, according to a trial published in The Journal of The American Medical Association (JAMA). The results also suggest that EarliPoint can be used as a proxy to accurately measure the child’s level of language and nonverbal learning.
In another two studies with 719 kids and 370 kids whose diagnosis of autism were still to be confirmed, the tracking tech accurately discovered an autism diagnosis about 82% and 81% of the time, respectively, according to a paper published in JAMA Network Open. It accurately identified the non-autistic kids as not having autism about 90% and 82% of the time, respectively.
The new tool reliably collects data that can be compared with potentially autistic kids’ typically developing peers in a more simple, straightforward, objective, and speedy manner compared to current gold standard practice — which usually involves hour-long assessments in which experts collate information from parents’ subjective reports about their kid’s developmental history and behavior, and then observe the child themselves.
The new eye tracking assessment is especially helpful since families testing their kids for autism currently face long wait times from referral to diagnosis in the U.S. There aren’t enough expert clinicians to meet the need for children with concerns, current diagnostic tools are often inaccurate or take a long time to run, and some insurance companies require many steps for diagnosis.
Because of these hurdles, research suggests that as few as 17% of autistic children are diagnosed before age 3. The average delay for diagnosis after a first autism screening is more than two years, according to a May 2023 study.
Training a general pediatrician to diagnose autism decreases the patient bottleneck wait time for an assessment from 135 to 68 days, according to preliminary findings from July 2023. And EarliPoint can be used to do just that. “The goal is to improve efficiency, augment volumes and access, decrease cost, and achieve the same high level of quality,” says Klin, “but truly make diagnosis not an end in itself, but an efficient gateway to treatment.”
Who Should Get The Autism Eye Tracking Test?
The new tech isn’t a silver bullet for autism diagnosis — and it wasn’t designed to be a standalone tool.
The data from the two studies on EarliPoint comes from kids who already had been referred for autism and got the test to confirm or deny their diagnosis. “These are ones where the parents are concerned, and they’ve been waiting for diagnosis,” says Raymond Sturner, M.D., a professor of pediatrics at the Johns Hopkins University School of Medicine, who was not involved in this research.
This means EarliPoint is not intended for general screening. “We do not have data on screening all kids,” Sturner says. “This is not going to be something your pediatrician is going to have in the office for screening, nor should they.”
You’re likely to be referred for this type of test only after your child has already been seen by your pediatrician — who regularly keeps track of their development and searches for signs of autism anyway — and only then if they have concerns about your child’s development and behavior. You can keep an eye on your child for these behaviors too, and tell your family clinician about them. These concerns, according to the National Institutes of Health, include signs such as your child:
- Being in their own world a lot
- Having poor eye contact
- Not trying to attract their parent’s attention
- Not being able to explain what they need or want
- Showing atypical attachment to toys or objects
Why 16 to 30 Months Is a Tricky, Yet Crucial, Age For Diagnosing Autism
Diagnosing a child with autism before they turn 3 is exceptionally tricky. Yet an early autism diagnosis can make a big difference for your child.
“A child’s formative years are critical, filled with rapid cognitive, social, emotional and physical development,” Christopher J. Smith, Ph.D., chief science officer of Southwest Autism Research & Resource Center, said in a statement about EarliPoint. “Sometimes differences in development are so subtle that parents and pediatricians are hesitant to act until delays become more problematic.”
But early diagnosis ensures they an autistic child has access to the right education and care as soon as possible — and that they’re receiving interventions at a time when their brain is very plastic, malleable, and predisposed to positive change. In research from March 2023, kids who underwent a nine-month treatment plan at the age of 18 months did better in skill tests for social communication and self-help than kids who accessed the same treatment just nine months later.
How and Where To Access Eye-Tracking Autism Diagnostics
Since the FDA’s authorization, EarliPoint has already been used in clinic at the Marcus Autism Center, and three other centers are setting up contracts to use the assessment, according to NBC News.
“Conversations are underway with several children’s hospitals and big insurance networks,” Klin adds. His team envisions the tool to be adopted by clinical diagnostic and treatment centers, children’s hospitals, pediatricians’ offices, and maybe even more accessible and immediate points of care, like a family’s local CVS. “It takes a minimally trained technician to operate it reliably and less than one hour of training,” says Klin.
The researchers are also currently testing whether EarliPoint could help diagnose autism in 9-month-olds and older children, Klin says, as well as how kids diagnosed with autism with the aid of EarliPoint do in the long run.
Most importantly, the team also has trials underway to test the usefulness of this tool as a population-based screener for everybody to use — even kids who aren’t suspected to have autism. “We hope that these data and the availability of this tool can help answer profound challenges that currently exist in the healthcare system,” Klin says.