Tuesday, September 08, 2009

fMRI: a mixture between neurological or behavioural??

I just had this frightful thought that we could be fooled by the functional brain imaging data. Here is my line of reasoning. Let's assume we have a structural issue (which is in my mind pretty much settled by genetics and structural brain imaging). So your speech system has a tendency to jam. Let's say in 1 word out of a 10. Let's say that the system randomly happened to jam more often on the d-word. I am wondering how the brain reacts to this apparent (but random) pattern. It could well start learning to associate the d-words with jamming. And the next d-words act as a trigger and the brain expects jamming. This expectation could make a real jamming more likely or might trigger the associated behaviors to jamming. So when we put someone in the scanner and give him to read, he might stutter on d-words, and we see a difference in the scans to fluent people. However, it is not clear to me that the functional differences is due to real jamming or just due to a trigger of associations. Let me rephrase: the functional data might be a mixture between abnormal neurological activity and activity due to normal associative learning not present in fluent subjects. This would mean that interpretation of functional data is very very difficult, and might not reflect abnormal function due to structural issues.

6 comments:

Anonymous said...

Tom,

"I just had this frightful thought that ..."

It looks like you're learning. It looks like you're starting to see the light and the truth. Why should that frighten you?

Gustaf said...

Good point. Another issue with the brain scans is with structural images. How do you know if the typical structural differences occurring among stutterers are due to a genetic difference, or whether they are a result of stuttering? The collected evidence points to a genetic predisposition for stuttering, but MRI is hardly the best evidence, and we should be wary of researchers using MRI technology to support their own pet theories.

Rafael said...

"How do you know if the typical structural differences occurring among stutterers are due to a genetic difference, or whether they are a result of stuttering?"

Diffusion tensor imaging studies with children.

Gustaf said...

Possibly, if there were such studies. Spanish is not one of my languages, but I understand that the study mentioned targeted teens and young adults. But even if it targeted children, I still don't think it's good enough. Because assuming the typical difference occurs as a result of stuttering, it's reasonable to think it would be somewhat visible on an MRI scan as soon as stuttering has become habituated (which is probably years before the child is selected for an MRI scan).

Tom Weidig said...

Gustaf,

we also know that stuttering has a significant genetics component for many stutterers, though a structural abnormality by birth is likely.

Tom

Gustaf said...

I think our basic premise is the same. It's just that whenever I hear someone talking about MRI scans as evidence for a genetic component in stuttering, I like to hear the argument behind. It's interesting to see the structural and functional differences in stutterers, but I don't believe they help us tell whether stuttering is a genetic disorder. Twin studies is probably the best evidence for that, but it's also kind of indirect. I believe that the final answer will be found in our genome eventually, and that's where I'd like to see more research.