A few days ago Carnegie Mellon University imaging expert Marcel Just published a very interesting study with the heady title, Identifying Autism from Neural Representations of Social Interactions: Neurocognitive Markers of Autism.
A simpler title would be, How Brain Scans Recognize Autism, and How Scans Identify Common Emotions. That's a bit bold for a staid scientific journal, but it sums up what they are doing.
Look here for the actual study, which is available to the public: plosone.org
In this study, Dr. Just put 34 young adults in a sophisticated scanner and asked them to imagine the verbs: compliment, insult, adore, hate, hug, kick, encourage, and humiliate. He asked them to consider the verbs from the perspective of applying them to another, and of feeling that about themselves.
Half the subjects were on the autism spectrum, the other half were not. They were generally matched for age (mid-20s), IQ (bit above average), handedness, and other factors.
As they held those thoughts in their minds, he created a three-dimensional map of activity in their brains. When rendered as 3-D color imagery, the result was a set of visual patterns that represented particular words and their associated emotions.
The findings were remarkable, to say the least. When you look at the images printed in the article, they look very similar. But a computer analysis that broke the brain down into tens of thousands of little cubes called voxels told a very different story.
Dr. Just was able to separate the autistic from the non-autistic subjects just by comparing their responses to the verbs. The autistic people did not show much activation of the areas we’d associate with emotional response to the words. Yet they showed the same activation as the others to the logical meaning. At least that's what he thinks, given the current understanding of the function of the different brain regions. What he knows for sure is that the autistic and non-autistic activations patterns differed in a consistent way.
Dr. Just suggests the autistic subjects responded as if they were “observing the emotions from outside a window” which is essentially how Temple Grandin and others have described their own emotional responses. I’ve written similar descriptions as well.
He cautions us that this study only included 34 subjects, and a much larger study would be needed to draw population-wide conclusions. With that caveat . . .
This study promises to point the way to a non-invasive tool that can distinguish autistic brains through a brief session in an fMRI scanner. That alone would be a major breakthrough.
But there’s more . . .
Temple and many others have written about our autistic sense of “looking in at the world from outside,” and feeling like “an anthropologist on Mars.” For the first time, Dr. Just has shown a biological foundation for that feeling. And not just that . . . he has shown that a group of autistic people responded to those verbs in essentially the same manner.
I’d like to clarify one thing the study DOES NOT say. Dr. Just is not suggesting that the autistic people have any more or less ability to feel emotions, or to respond, than anyone else. Rather, he is showing that autistic and neurotypical people respond very differently when told to, “Imagine that you are insulted!”
I’ve written at length about that very thing from the perspective of life experience, and I am just staggered to read his exploration of the underlying biological foundation—the voxels of the brain
And the final thing, which some of you will find chilling . . .
Dr. Just has shown that he can read emotions from patterns of activated voxels in those 3-D brain maps. He can look at a pattern and tell you which of the verbs (emotions) the subject was imagining or feeling at that moment.
The day of wordless conversation between computers and us is coming fast.
There is really a lot to think about in this study. This is a brilliant piece of work, and it’s freely available. I encourage you to read it, think through what it may mean, and share your thoughts. In my opinion, this is a very big deal, and I congratulate the CMU research team for their fine work.
It would be very interesting to see this study expanded with a bigger group, including the whole IQ and affect range on the autism spectrum. Is this pattern similarity shared by all of us? I wonder . . .
Disclaimer: I know Dr. Just, and have visited his lab on several occasions. I have discussed this study and others with him outside of the article cited. I'm a strong supporter of his work, and of other work like this, because I believe it can lead to better understanding and ultimately better supports for people on the spectrum. That said, I think the study speaks for itself and my personal connection just made it all the more meaningful to me. I have no other association with Dr. Just and his lab or university.
John Elder Robison is an autistic adult and advocate for people with neurological differences. He's the author of Look Me in the Eye, Be Different, Raising Cubby, and the forthcoming Switched On. He serves on the Interagency Autism Coordinating Committee of the US Dept of Healthand Human Services and many other autism-related boards. He's co-founder of the TCS Auto Program (A school for teens with developmental challenges) and he’s Neurodiversity Scholar in Residence at the College of William & Mary in Williamsburg, Virginia. The opinions expressed here are his own.
Teaser Image: Copyright John Elder Robison