Research uses MRI to identify attention deficit hyperactivity disorder

When magnetic resonance imaging, or MRI, was first introduced as a diagnostic tool in the 1990s, a new world was opened to researchers as the scans provided a novel way to look at the inner workings of the human body. Now, researchers at Yale University have found a way to use the MRI to see even further into the human psyche to identify attention problems.

In a study recently published in Nature Neuroscience, the Yale team presented evidence that they could correctly identify adults and children with a range of attentional issues through MRI scans that focused on connectivity between the regions of the brain. Monica Rosenberg, Ph.D. candidate in psychology at Yale, spoke with New England Psychologist’s Catherine Robertson Souter about the findings and what they could mean to the treatment of ADHD and to the field of psychology as a whole.

Q:  When you talk about the work that you did with adults and children in this study, you mention the idea of brain connectivity and how everyone has a network pattern as unique as a fingerprint. Can you give us some background so that we can understand the impact of the attention deficit findings?

A:  Functional connectivity is different than an actual structural connection. What we mean is that the activity is synchronous between regions; with two regions that have a strong functional connectivity, the activity in each would go up and down at the same time. The co-lead author on this study, Emily Finn, also a graduate student here at Yale, was a co-lead author on another study published in October showing that every person has a unique pattern of functional connectivity.

So once Emily and her lab got this finding – we were encouraged to see if we could find information about people from these patterns of connectivity.

Q: Tell us about the study.

A:  The first thing we did was to scan some adults here at Yale in an MRI scanner and collect brain data as they were performing a challenging attention task. We looked to see whether there were brain connections that related to how well they performed on the task. When we look at the results, it gives us a really complicated brain network that predicts attention.

We also collected brain data when they were lying still looking at a simple cross on the screen.

We found that we could get significant predictions of how well a person was going to do on the task not only from their pattern of brain connectivity while they were performing the task but when they weren’t doing anything at all. This suggested we could take any person and learn how they will perform on an attention task that we will later give them.

The next question to answer was whether we were measuring brain connections that were specific to the task we were giving or whether we were measuring something more general about attentional ability.

Luckily, there was a data set publicly available collected in China of children and adolescents resting in a scanner and it included a measure of their ADHD symptoms. We tried to predict for each of these children, by looking at the resting scan data, how they would perform if we were to hypothetically give them our attention task.

We found that the children we predicted would do well had low ADHD symptoms and the opposite. This suggested to us that the brain networks that we were identifying were getting at something that was generalizable across these two very different populations and very different measures of sustained attention.

Q:  What did you learn about ADHD and diagnoses?

A:  Our predictions are not perfect so I would not base a clinical diagnosis on them. Instead, I would emphasize the continuous nature of our prediction. We are not just predicting ADHD but something a little more nuanced which is how severe an attention deficit someone has. We are predicting on a continuous scale where they fall relative to other people.

Q:  Were the children in the study on ADHD medication?

A:  Some of children who had ADHD diagnoses had previously been on medication but they were not actively on medication at the time of the scan.

Q:  Which leads to a question about brain connectivity. Does it change if you are on medication or taking drugs or drunk?

A:  That is something we are actively exploring. It would be exciting to find that these meds work by affecting brain connectivity. We are hoping to be able to have some more answers on that in the near future.  
Q:  Is this area what you are working on now?

A:  It is really hard to collect MRI data from hundreds of kids, so we are pursuing collaborations.

We are hoping to find a group that has maybe analyzed the data with a different question in mind or with a different technique and coming in and seeing if our method could shed some light on how these medications are working and on what effect they have.

In addition to investigating the effects of ADHD medications, we have two on-going projects in the lab – one is looking at fluctuations in attention within a single person. The other is an on-going study trying to see whether we can affect a person’s brain connectivity to have a positive impact on their attention later.

Q:  How do you affect someone’s brain connections?

A:  We are not sure we can do it. One thing we are trying is giving feedback on how strong a particular brain network is and maybe over time, we can train people to express it more strongly.

Q:  Like biofeedback?

A:  Yes, it is similar. There is a new technique which allows us to give feedback according to the MRI signal instead of being based on heart rate.

Q:  Are there other things that this can predict? Alzheimer’s, schizophrenia, depression?

A:  In the paper published in October they did show that within one sample of individuals you could predict fluid intelligence using this same method. Moving forward, we are interested in seeing if we can predict what is to come, to see if someone will go on to develop attention deficit or other things. If we were to acquire some longitude data we are optimistic that we would be able to predict future changes in symptoms or abilities. That is our long-term goal.

Q:  With ADHD, do you think this could change treatment?

A:  That is the hope. Now that we have identified the networks involved in ADHD, can we target those networks and can we find some way to strengthen those brain connections and will that have an effect on life outcomes and school performance?

Q:  Why is this discovery important?
 
A: One reason is just because of generalizability of the methods. We have demonstrated that we can take any person, an adult in New England or a child in Beijing, and have them rest in a scanner for as little as eight minutes and can predict something really complicated about them.

The most exciting thing, I think, we have done here is show that there are patterns of connectivity that scale with symptoms. ADHD is not an all-or-nothing disorder. It is not that we just described one small sample we collected but we seem to be describing something about the population at large. I don’t think we have identified the most generalizable networks yet but we are moving in that direction.