Imaging Sheds Light on Autism


by Faith Hickman Brynie

September, 2007

Many children and adults with autism lack, to a greater or lesser degree, the innate ability to grasp and interpret the meanings of gestures, body movements, tone of voice, and facial expressions. Now researchers are using brain-imaging technologies to discover why the brains of children with autism fail to process social inputs.

One powerful tool for such research is functional magnetic resonance imaging, or fMRI. The MRI scanner uses a large magnet to deliver magnetic pulses that induce a tiny radio signal in the part of the body being scanned. Computers detect the signals and generate images from them.

With fMRI, fast scanning techniques allow investigators to trace the flow of blood in the brain, which increases in areas where the brain is working harder. By comparing images when the brain is “at work” or “at rest,” investigators can see which regions become more active during mental tasks.

Scientists are using fMRI to study the brain’s mirror neuron system in children with autism. Mirror neurons fire when a person performs an action with a goal in mind or watches someone else perform an action. Neuroscientists think that mirror neurons allow one person to understand the actions, intentions and emotions of another.

Mirella Dapretto, an associate professor of psychiatry and biobehavioral sciences at UCLA, and her team have used fMRI to study the activity of mirror neurons in the brains of 10 children with autism. She also studied 10 normally developing children. First she used behavioral measures to assess the children’s ability to imitate and empathize with others. Then she captured fMRI images of brain activity as the children viewed and imitated pictures of faces depicting anger, fear, happiness or sadness.

Those children who best imitated social behaviors or best empathized with others had more brain activity in the brain’s right inferior frontal gyrus. Those who showed the greatest deficits in imitation and empathy had less activity in that region, which is known to be an important area of mirror neuron activity.

 “Our findings suggest that a dysfunctional mirror neuron system may underlie the social deficits observed in autism,” Dapretto says.

In a separate but related study, UCLA clinical neuropsychologist Susan Bookheimer and her team used fMRI to find out why children with autism cannot interpret facial expressions the way other children can.

Bookheimer had 16 normally developing children and 16 children with autism look at photographs of people showing angry, fearful, happy and neutral expressions. The researchers digitally manipulated the faces into two different forms. In some pictures, the eyes gazed straight ahead. In others, the gaze was averted; the eyes looked sideways.

Children who do not have autism clearly distinguished between direct and averted gazes. Direct-gaze expressions activated several brain areas that have mirror neurons, especially an area called ventrolateral prefrontal cortex. When these children viewed the averted-gaze pictures, this area quieted down.

The results were different among children with autism. They showed little or no activity in the ventrolateral prefrontal cortex, regardless of the subjects’ gaze or expressions in the photos. They also showed no evidence that they processed the emotions, whether the eyes were direct or averted.

“This part of the brain (the ventrolateral prefrontal cortex) helps us discern the meaning and significance of what another person is thinking,” says graduate student Mari Davies, a member of Bookheimer’s team. Davies thinks inactivity in the VLPFC may explain why children with autism do not recognize visual cues and do not process the emotional content of conversations.

The goals of fMRI imaging research go beyond localizing differences in brain activity. “I think that this research is crucial to design therapies that are the most appropriate to autism,” says Nouchine Hadjikhani, an associate professor in radiology at Harvard Medical School.

“Understanding the underlying differences in brain processing is the first indispensable step,” she adds. “One needs to be sure to address the underlying cause, not just the symptoms.”

If dysfunctions in the mirror neuron system are part of autism, Hadjikhani thinks, then those who work with children who have the disorder can take advantage of brain plasticity to train children in skills such as imitation and empathy.

Dapretto reports progress in designing and testing one such training protocol. She gave children with autism specific instructions to pay attention to facial expressions as they viewed a series of pictures. Functional MRI images revealed increased activity in the medial prefrontal cortex, a part of the mirror neuron network.

“This is a very positive thing,” Dapretto says. “These findings have implications for future interventions. They suggest that you could train the autistic brain to make use of the information conveyed by the human face and voice to successfully navigate social interactions.”