Screen shot from Astropolis, a video game designed to test a range of domains in children with autism and their siblings.

CORNELL (US)—A custom-built suite of science-fiction-themed video games may help researchers uncover clues to the chaotic nature of autism.

Matthew Belmonte, assistant professor of human development at Cornell University, worked with collaborators in computer science to develop the video game suite called Astropolis.

Unlike much of the current research on autism, which isolates and tests a single domain, Belmonte designed the user-friendly games with embedded tasks that test users—children with autism or Asperger syndrome ages 10 to 15, along with their unaffected siblings—across multiple domains.

“Autism has been characterized as a fundamental perceptual abnormality; it’s been characterized as a fundamental attentional abnormality; it’s been characterized as a failure of theory-of-mind,” he says. “We each have our individual pet theories, and we each—me included—have designed experiments within these narrow theoretical apertures to confirm or refute hypotheses that are stated along our single tracks.”

In this experiment, he hopes to find links between social and nonsocial theories; and between behavioral and physiological data.

He also hopes to sidestep a confounding factor common to autism research.

“When we look at people with autism in a lab, it’s not clear that we’re testing them under naturalistic conditions, because one of the hallmarks of autism is very high levels of anxiety—anxiety with new people, and anxiety with new places,” he explains.

As a consequence, researchers often can’t be sure if their findings are due to autism itself or simply heightened anxiety.

Belmonte gets around that by giving the games to subjects on laptops to take home and play at their own pace. While the subjects grapple with asteroids, pilot spacecrafts, intercept pirates, and salvage hidden cargo, the computer records how rapidly they shift attention and engage or inhibit motor responses; how well they perceive coherent motion; and whether they can intuit the motivations of other characters.

After the subjects are comfortable with the game on their own, Belmonte uses electroencephalography (EEG) in the lab to measure patterns in neural connectivity as they play.

Earlier research has shown that while autistic children and their unaffected siblings share some physiological traits—including frontal lobes that are slower to activate— the overall patterns of neural connectivity are weaker in autistic children compared with their unaffected siblings. “We’re going to be looking at that in much larger numbers now using EEG,” Belmonte says.

The project’s broad, multivariate scope is part of an emerging trend in autism research, he notes. “The field is starting to recognize that you need to do this correlative cross-domain work to figure out what’s developmentally related.”

In the 20th century, neurobiologists had success tackling problems tied to a circumscribed brain region, he adds. “We’re now left with the hard problems, in which the deficit is not so much in one localized region, but in the network and the way brain regions talk to each other to implement complex functions.”

The research is supported by Autism Speaks and by the National Science Foundation through the American Recovery and Reinvestment Act.

Cornell University news: