Fast ForWord helps children with ASD improve their retention and memory skills.

With autism diagnoses on the rise, autism awareness organizations are banding together to increase early detection and education practices for students with the disorder.

According to the Center for Disease Control, roughly 1 percent of all children in the United States were diagnosed with Autism Spectrum Disorders (ASD) in 2009—an increase from the 1 in 150 children diagnosed in 2007.

Hoping children on the autism spectrum might be diagnosed at an earlier age, thereby providing more treatment options, the National Science Foundation has awarded $10 million to a conglomerate of research universities, headed by Georgia Tech, to develop a computer screening system for earlier autism diagnoses.

“The idea is to try to identify abnormalities in very young childhood development sooner than is currently possible and to try to do this for more children than it’s possible now,” said Ephraim Glinert, the program officer for Georgia Tech and its partner institutions.

Early autism detection has taken a front seat as autism awareness increases across the nation.

“If we can detect real problems very early, then it might be possible to treat them … to either eliminate the risk or reduce the severity of whatever problems there may be,” Glinert said. He said that as children grow older, it becomes more difficult to treat the symptoms of ASD.

The program being developed by Glinert and his research team will record everything a child does over a predetermined period of time using cameras and microphones, and then a computer system will process the data. The system will flag any sign of troubling behavior, at which point a doctor can examine a particular case.

“It’s impossible to even train enough clinicians to give everyone the attention and the quality [of] medical care that they really want,” Glinert said. He said he hopes computers will take the brunt of the burden, so that physicians will be able to focus on actual cases of ASD and not false alarms. Researchers also are hoping the system will pick up cues far earlier than a doctor would be able to.

“I think they’d like to get it down to before a child is even 2; very early in their development, when it’s too soon for a clinician to normally be able to identify what the problems are,” Glinert said. “If we could find a way to get a heads up on it from the computer that there may be an issue, then you’d of course bring in the live clinicians for those cases where it seems important.”