Seeing kids getting excited about cartoons is hardly cause for alarm, but when David Grelotti, a former research assistant at the Yale Child Study Center, observed that autistic children knew more about cartoons than about people, he took note.

“I thought it was interesting that they knew all the names of Pokemon and Digimon, but they didn’t know my name after three days,” Grelotti said.

He worked with Robert Schultz, a professor at the Child Study Center, using images of Digimon characters in a study of autistic children published earlier this month. Their results indicated that while autistic children cannot recognize faces as healthy people can, they have normally functioning brain tissue in the fusiform gyrus — the area of the brain specialized for facial recognition.

Work done at Yale by Isabel Gauthier, currently a professor at Vanderbilt University, found that the fusiform face area is moldable. It activates when a person with extensive familiarity with a certain class of object differentiates between two objects of that class, such as a car enthusiast noticing the difference between a 1964 Buick Skylark and 1963 Pontiac Tempest.

Gauthier created computer blobs called Greebles that had certain regular features. A person who was not acquainted with Greebles took a significant period of time to differentiate between two Greebles, whereas an expert could spot the difference more quickly. Schultz said Gauthier trained Yale undergraduates to be Greeble experts and scanned their brains using functional magnetic resonance imaging.

“Lo and behold, when she showed them Greebles, there was activation in this face area,” he said.

Autistic individuals are not face experts and do not show activation in the fusiform gyrus when they see faces, Schultz said. Under fMRI scans, the fusiform areas of autistic children showed more activity in response to furniture than to faces.

Schultz’s research was supported by the work of Ami Klin, also a professor at the Child Study Center. Klin showed autistic children videos of caregivers approaching children and found that autistic children do not have the usual interest in others’ eyes.

The fusiform only responds to objects of a person’s expertise. However, a person can only become an expert at something if he or she is interested in a class of objects and pays attention to them. The amygdala is an area of the brain involved in emotional processes, Schultz said. It detects things that are useful for a person’s survival, such as the expression on someone else’s face. If the amygdala is dysfunctional, then a person can never become a face expert, and the fusiform face area will not respond to faces.

“We know the fusiform is underactive in autism,” Shultz said. “We don’t know whether this causes autistic symptoms or whether it results from being autistic.”

Schultz said he believes the underactivity results from autism. Autistic people do not focus on faces, so they never become face experts, and they do not develop the fusiform face area, he said.

In order to test this theory, the researchers needed something in which many autistic children were expert. They found it in Digimon, a cartoon about digital monsters that has spawned a franchise. One autistic patient who spent hours every day watching the show or playing Digimon games influenced Grelotti’s choice.

“With normal people, faces are special. With this kid, Digimon was special,” he said.

The researchers mapped the child’s brain activity while showing him pictures of Digimon characters. The fusiform area showed normal activation in response to this abnormal stimulus.

“His brain, when he looked at something he liked, acted like our brain does normally when we’re looking at social stuff, at least in the fusiform gyrus and amygdala,” Grelotti said.

The experiment provided evidence for Schultz’s hypothesis, showing no inherent dysfunction in the fusiform gyrus.

Schultz said if an autistic person is emotionally engaged with an object, that person can become an expert in it. If a person becomes an expert in recognizing faces, that person may be able to glean social information from faces.

“The promise of Digimon is … if we just know how to engage the brain circuits, the fundamental tissues and pathways are all there,” Schultz said. “How do you get kids with autism to care about faces?”

To answer his own question, Schultz has created a series of games called Let’s Face It. One of the games resembles the popular computer game Snood, but, instead of matching cartoon heads, the player has to match up faces. At the lowest level, the player matches identical pictures, and at the higher levels the player has to group pictures of the same person photographed from different angles.

The games can be played in two modes — identity mode and expression mode. Using both modes, Schultz said children can train themselves to identify people and recognize social information from a face.

So far, only one child has trained on Let’s Face It, and that child showed increased amygdala activity in response to faces after training. Schultz said this signals that the child now understands faces to be objects of interest — the first step towards social interaction. Then the child can develop skill in recognizing identity and expression.

Schultz said he does not know whether the failure of the amygdala to respond to faces is a large or minor part of autism. He plans to continue training autistic children with Let’s Face It to observe whether the program is successful in helping children overcome some symptoms of autism.

In his next study, Klin will follow a group of babies from birth but will only study children with autistic siblings. He predicts five to ten percent of these infants will develop autism. With this method, Klin said he will be able to examine autism in its earliest stages and observe how symptoms develop from birth.