Diffusion-based magnetic resonance images are sensitive to water motion in the brain. Diffusion tensor imaging (DTI) provides information on microarchitecture of the brain's white matter that reflects changes in axonal myelination processes and, therefore, the integrity of organized tissue microstructures.

“The part that we can pick up is the water motion that is in the sulci next to the gray matter [in the cerebrospinal fluid]. This is very easy to measure,” said Manzar Ashtari, Ph.D., a senior neuroscientist in the radiology department of the Children's Hospital of Philadelphia.

Dr. Ashtari used a DTI measure called the apparent diffusion coefficient (ADC) to measure cerebrospinal fluid (CSF) changes in the sulci as a surrogate for gray matter changes. The ADC is an average measure of the diffusion of water in all directions in a single voxel. ADC is greatest in CSF and least in coherent, healthy white matter. The idea that ADC could be a surrogate marker is based on the observations that cortical brain atrophy is associated with a corresponding increase in sulcal and ventricular CSF, meaning that CSF may be considered a tracer for detecting cortical gray matter reductions.

ADC rises with increasing CSF volume, which may represent gray matter volume loss. “We are basically saying that if you look at the water motion in the cerebrospinal fluid that is in the sulci, whatever changes that happen to the CSF could be a surrogate for the gray matter right next to it,” said Dr. Ashtari.

Dr. Ashtari used the new technique to look for potentially subtle differences in gray matter volume in high-functioning autism or Asperger syndrome patients. The autism imaging literature on white, gray, and whole-brain volumes is very confusing in part because imaging findings vary by age, she said.

The brains of autistic children appear not to go through normal growth processes, as compared with the neuronal development of healthy children. Although autistic children start with larger brain volumes, eventually normal children catch up with and pass their autistic counterparts in terms of gray or white matter volume.

In her study, Dr. Ashtari looked only at preadolescent boys (average age 11 years for both controls and autistic children). Autism most often affects boys, by a 4:1 ratio. She specifically looked at a preadolescent population because previous work of hers showed a big spurt of myelination during adolescence in healthy participants (Neuroimage 2007;35:501–10).

Quite unexpectedly, “We found increased gray matter [decreased ADC values] in several areas of the cortex,” in autistic preadolescents. “I don't want this study to be generalized. I'm not claiming that if you look at [autistic] adults, this is what you'll see,” said Dr. Ashtari.

The areas of abnormality occurred most often in the parietal lobe, which has a strong connection with the prefrontal region. “I concluded that I'm looking at mirror neuron abnormalities in autistic kids,” she said. Mirror neurons are thought to fire both when an animal acts and experiences an emotion or sensation and when the animal observes the same action, emotions, and sensations in others. They mirror the behavior of another animal, as though the observer were acting itself. These neurons have been directly observed in primates and are believed to exist in humans.

Next, Dr. Ashtari looked for possible correlations between increased gray matter and specific autism behaviors. To do this she used the Autism Diagnostic Interview and the Autism Diagnostic Observation Schedule.

They also measured the children's IQs.

She plotted the inverse of the ADC (which represents gray matter changes) against the measures of stereotyped behavior/restricted interest. She found that the greater the gray matter values, the poorer the performance (correlation factor of 0.6; P value of .04). She also plotted gray matter changes against measures of abnormalities in social interactions. Greater gray matter volume trended toward greater abnormalities in social interaction (correlation factor 0.5; P .08). Next, she plotted gray matter volume against IQ for both normal children and those with autism. She found that normal children have increased IQ with increased gray matter but there was no correlation in children with autism and IQ. In fact, the trend was the reverse; with increased gray matter, children with autism showed a decreased IQ.

“The conclusion I made was that this increased gray matter is nonfunctioning gray matter,” said Dr. Ashtari.

“There was one other area of abnormality that we found and that was deep inside, not really on the cortex … basically around the right amygdala area,” said Dr. Ashtari. The amygdala is a center for emotional processing. The researchers found decreased gray matter in the right amygdala of children with autism.

70043_fx1.sml

Apparent diffusion coefficient-based morphometry revealed clusters of increased gray matter (highlighted) in the right and left parietal cortex of an autistic boy. Courtesy Manzar Ashtari, Ph.D.

Image of the Month