Developing a Test for Autism

A revolutionary new test takes the guesswork out of confirming autism.

Researchers at Harvard have found a new way to use brain scans to confirm the presence of autism. Rather than look for structural abnormalities of the brain, these researchers decided to look at the circuitry that links these structures. The test makes use of preexisting technology in the form of MRI brain scans.

Until now, there was no subjective way to diagnose autism: no clinical tests perform that would prove or disprove autism. Autism could only be presumed by observing the subject’s autistic behavior.

Gold Standard

Previous efforts to find a biological test for autism have been a failure. Researchers looking for the gold standard of a biological autism test had focused in the main on viewing the various structures of the brain, looking for differences between the typical and the autistic brain. But this search yielded no fruit.

Harvard Medical School’s Nicholas Lange and his team decided to look for anomalies in the circuitry that links the brain structures rather than at the functional brain structures themselves. “The brain may be okay, the parts that do the work may be okay,” said Lange. “But the wiring—the ‘cables’ between the points in the brain, one to another –may be disrupted in some way.”

Harvard Medical School

The circuitry of the brain is also called “white matter” and can be seen with magnetic resonance imaging scans (MRI). Lange applied a specific technique called diffusion tensor imaging that can show to what extent the brain’s circuits are organized.

Uncooked Spaghetti

Lange says the brain’s circuitry is not unlike a package of uncooked spaghetti, all neat and organized within its packaging. After the spaghetti is cooked, however, it’s just a disorganized tangle. The circuitry of the brain can be very organized like the neat packages of spaghetti on a grocer’s shelf, or it can resemble a tangle of pasta in a colander.

With the help of MRI scans, the researchers were able to assess the organization of these connections in the two parts of the brain dedicated to language and social functioning. According to the results of their work, just published in the journal Autism Research, the scientists were able to use these scans to tell who had autism with 94% accuracy. A later study performed to confirm these findings had an even greater success rate, making this the most accurate biological test for autism, yet.

Lange believes the application of this research is most helpful for other researchers at this point. Further studies need be done to prove the worth of this diagnostic tool; though Lange feels such scans might be helpful in confirming autism at an earlier stage than is possible at present, making early intervention a possibility. At present, autism cannot be diagnosed before the age of three.