Formula for Earlier Autism Detection on the Anvil

by Rajashri on Aug 20 2008 4:47 PM

3-D imaging to study correlations in the facial features and brain structures of children with autism spectrum disorder (ASD) is being used by University of Missouri researchers. They believe can help them develop a formula to detect the disorder early.

"When you compare the faces and head shapes of children with specific types of autism to other children, it is obvious there are variations," said Ye Duan, assistant computer science professor in the MU College of Engineering.

"Currently, autism diagnosis is purely behaviour based and doctors use tape measurements to check for facial and brain dissimilarities.

"We are developing a quantitative method that will accurately measure these differences and allow for earlier, more precise detection of specific types of the disorder.

"Once we have created a formula, we can pre-screen children by performing a quick, non-invasive scan of each child's face and brain to check for abnormalities. Early detection is crucial in treating children and preparing families," Duan added.

Kevin Karsch, a research assistant in Duan's computer graphics lab, MU senior and Goldwater Scholar, said: "Instead of looking at brain structures slice-by-slice in an MRI (magnetic resonance image), we developed tools to create 3-D representations of the structures in order to visualize and make comparisons."

"Using the 3-D representations, we are comparing the brain structures of autistic children to those of non-autistic children; no one has ever done that," he added.

Duan's facial and brain imaging work will focus on two ASD subgroups hypothetically identified by Judith Miles, the William and Nancy Thompson Endowed Chair in Child Health at the Thompson Center.

She has observed and distinguished children with a tendency toward more physical and brain abnormalities and smaller heads as having complex autism. Only about 20 percent of affected children fit this subset.

The other 80 percent are classified as having essential autism. Miles also has identified physical similarities in facial structure and increased cranium size among those in the essential group and has speculated that the traits may be related to brain abnormalities.

To address the pressing questions about brain development and function in autism, a dozen University of Missouri researchers from a variety of disciplines formed The Autism Neuroscience Research Group (ANRG).

"Our research group is exploiting a tremendous overlap of knowledge from many disciplines. Ye's work will be a pivotal part of ANRG's grand scheme. To solve these problems, it takes an entire community of specialists working together," Miles said.