Can Measuring Blinking Help Detect Autism Early?

It’s estimated that humans blink about 10 to 25 times a minute. The primary purpose of blinking is to keep the eye moist. But its purpose goes beyond that simple function. When a person is looking at something they find especially interesting or engaging, such as when they are playing a video game, their blink rate drops to fewer than five blinks per minute. Why does this happen?

By reducing how often we blink, we can maximize the amount of visual information that is taken in and processed. When we find something we are looking at engaging, remarkably, we unconsciously suppress our blinking to increase the visual flow of information to our brain. The next time you are at dinner with a friend or colleague, notice how often they blink when you are talking with them. It could give you a clue whether they find your conversation interesting!

Dopamine

One theory is that blinking is linked to the part of our brain that is activated by a rewarding experience. Some evidence suggests that blinking is regulated by dopamine, a transmitter that is associated with the experience of reward. It makes sense that when we find something rewarding, our brains automatically increase our ability to pay attention to the thing we like. Dopamine has also been linked to cognitive flexibility, the ability to rapidly change one’s attention and perception from one thing to another.

Blink rate has been used by scientists to track how people allocate their attention when they are engaged in a variety of activities. For example, blink rate has been found to drop when someone watches an emotionally evocative movie. Spontaneous blink rate while completing a task that requires focus and memory can predict how accurately the person is able to complete the task. Blink rate has been used to study what factors interfere with a pilot’s ability to sustain their attention and its effect on their flight performance.

Detecting Neurological Differences

Researchers are now exploring whether blink rate can help detect neurological differences and disorders, such as Parkinson’s disease and autism. Parkinson’s disease is a brain disorder related to changes in dopamine levels that cause difficulties in coordination and balance. Medications can help a person with Parkinson’s disease, but understanding how different doses of medication are affecting the person’s dopamine levels can be challenging. Because blink rate is correlated with dopamine, doctors have found that monitoring a Parkinson’s disease patient’s blink rate can offer an easier way of tracking whether a medication is working.

Recently, our team at Duke University published a study that suggests that blink rate also could offer a way of detecting early signs of autism. Using an app downloaded on a tablet, we showed hundreds of toddlers a set of brief, engaging movies. Some of these toddlers were subsequently diagnosed with autism. The movies varied in their content. Some movies were of people, such as a woman singing nursery rhymes. Other movies were of objects, such as a spinning top. While the toddlers watched the movies, the camera in the tablet recorded the children’s faces. A team of engineers at Duke then used a technology called computer vision to measure how often the toddlers blinked while they watched the movies.

We discovered that the neurotypical toddlers suppressed their blinking more when they were watching movies of people, whereas the autistic toddlers showed the same rate of blinking regardless of whether they were watching people or objects. Autism is a condition in which the child often finds the world of objects much more interesting than the world of people. Using this short snippet of behavior (about six minutes), we could distinguish children who were later diagnosed with autism with 82 percent accuracy by examining whether they tended to face the screen during the social movies combined with their blink rate. These objective measures of behavior, called “biomarkers,” may eventually help us do a better job of detecting autism in young children.

If you take me up on my suggestion to observe your friend’s blink rate the next time you are having a conversation, you might discover that it’s hard to do. I bet it’s so hard that you would have to focus your attention so much that your own blink rate would likely plummet. There are many subtle changes in our behavior that can reveal a lot about what we find interesting—not only blink rate but also things such as how dilated our pupils are and how often we move our head when we watch a movie. It turns out that a computer can do a much better job of detecting these subtle differences in our behavior than the human eye. In the future, we likely will rely increasingly more on technologies such as computer vision to gain a deeper understanding of human behavior.