Brain Imaging Studies in Autism Spectrum Disorders

Helen Tager-Flusberg, Ph.D., Boston University School of Medicine

For the past ten years our research laboratory, now located at the Boston University School of Medicine, has been conducting brain imaging studies on children and adults with autism or Asperger Syndrome (referred to here as ASD) in collaboration with colleagues at Massachusetts General Hospital using magnetic resonance imaging, or MRI. When we began there were very few published studies that had used this technology to investigate the brain structure and function in people with ASD; however in recent years there has been an huge increase in the literature in this area as more research groups take advantage of these exciting methods that allow us to investigate the brain in exquisite detail, in a painless non-invasive way. We have learned a good deal about some of the differences in the ways in which brains develop in children with ASD and how they function when processing different kinds of information. In this article I summarize some of the key findings, including work conducted by our group as well as others in the field.

For us to collect structural images depicting the anatomy of the brain using MRI, a person lies horizontally in the bore of a large “magnet” which produces a strong magnetic field. The procedure can take up to 20 minutes and the person must remain completely still while hearing loud noises generated by the magnet. For these reasons, MRI has more limited use with younger children, or individuals with high levels of anxiety, hyperactivity, or sensitivity to noise. In our lab, we spend considerable time acclimating our participants in a “mock” scanner to help prepare them for the challenges of lying in the magnet. When we collect structural scans, our participants are able to watch and listen to their favorite movies, which help them to pass the time despite the discomfort of the MRI scanner. Functional MRI (fMRI) images are collected in the same way; however the purpose of this method is to identify brain areas that are most ‘active’ when processing specific types of stimuli, such as faces, speech, or motor movements.

We have been collecting structural MRI scans from children and adults with ASD ranging in age from 4 to 24 for quite a number of years now. Much of our work has focused on the key areas of the brain associated with language that are typically located in the left hemisphere. (The outer region of the brain–the cortex–is divided into two halves, or hemispheres, connected by a bundle of nerve fibers called the corpus callosum). This association between the left hemisphere and language leads to asymmetry in the volumes of these “language” regions. One consistent finding from our work, across several different groups of participants, is that we do not find the same asymmetry patterns in ASD. Although there is considerable individual variability, asymmetry is less pronounced in ASD than among typical age-matched controls. The degree of reduced asymmetry is associated with both language skills and with severity of autism symptoms especially in younger children, according to our most recent findings.

In some of our other investigations we have examined the thickness of cortical areas. We found that in the right hemisphere, the areas associated with social functioning and imitation–part of the so-called “mirror neuron system”—have reduced thickness, and that this reduction is also associated with severity of social symptoms in ASD.

There are also some important brain areas associated with social functioning that lie below the cortex, in particular the amygdala. This region shows atypical patterns of development in ASD; by adolescence, the smaller size of the amygdala is also associated with severity of social symptoms.

Across many different research groups, including ours, one of the most replicated findings is that brain volume is often larger than average in individuals with ASD, particularly in children. It is thought that this enlargement takes place during the first few years of life. It is not clear what events lead to this increase in size, but it seems to be primarily in the frontal areas of the cortex, areas that include some of the main regions important for language and social information processing. The white matter of the brain, which is important for making connections across different cells and regions, seems to be responsible for the increase in brain volume. At the same time, the size of the corpus callosum is reduced in ASD. These findings have led to the important theory that in ASD there are differences in how well different areas of the brain are connected: areas that are far apart are not as well connected, which affects how efficiently complex information is processed, especially information that requires integration across visual, spatial, linguistic and cognitive skills.

Our functional MRI studies also focus on language and social processing. In our language studies, we find that even in high functioning verbal children and adults with ASD, there are differences in brain activation patterns. Like typical controls, people with ASD process language in the left hemisphere, in the same key areas of the brain. However, the patterns of activation are distributed differently, suggesting that in ASD the brain works differently, and that the language areas are not as well integrated or connected as they are in neurotypical people; such connections are critical for complex language processing, including higher level discourse.

Many studies have investigated how the brain processes faces, a key social stimulus, in ASD. Early studies found the main area for face processing, called the FFA (face fusiform area), was not activated in ASD. However, in our studies we required our participants to look directly at the center of the face while we collected the brain images. This difference in procedure showed that people with ASD do show typical FFA activation when they are attending to the face. Our collaborators at the University of Wisconsin went one step further, measuring where their participants looked, and which areas of the brain were activated, while they were judging emotional facial expressions. They found a direct relationship between FFA activation and amount of time spent looking at the eye region of the faces in their participants with ASD. Across both these studies we also found differences in brain activation in other regions associated with social processing, including the amygdala and the mirror neuron system.

We are continuing to conduct both structural and functional brain imaging studies at our research center at Boston University. To learn more about our work, visit our website We are always seeking new children and adults who are interested in participating. Participants are reimbursed for travel and parking costs, and compensated for their time. The future of brain imaging research in the field of ASD holds much promise; yet we still have a great deal to learn about the brain bases of the social, communicative and behavioral patterns associated with autism and Asperger Syndrome. We hope you will join us in the work that we are doing by being a part of our research program!