Q: How do you use MRI to study brain function?
A: We make MRI movies of the "blushing brain", from which we calculate "activation maps".
Q: It blushes?
A: Changes in neuronal activity are accompanied by changes in blood flow and blood oxygenation (as has been known since the 19th century).
Q: You make movies?
A: We image the whole brain every two seconds (30 frames per minute).
Q: How do you calculate that "activation map"?
A: Generally speaking, we use the MRI movie to make a picture of where the "brain blushes" in time with stimuli or responses. Those colored blobs (that we overlay, photoshop-style, over a greyscale image of brain anatomy) show where the timing of the MRI signals agrees with the timing of the stimuli or behavior.
Q: What do those maps look like? What do they show?
A: They generally show a collection of brain regions, that do different things (principle of segregation), which work together (principle of integration).
Here's an example: A right-handed adult (who gave written informed consent to participate in IRB-approved research) was scanned for several minutes, during which they alternated, for thirty second periods, between holding still, and sequentially tapping the fingers of their right (dominant) hand – the thumb to each of the opposing fingers, in turn. Looking at just one "slice" – not even the whole brain – you can see four different regions that serve four different roles. Our understanding of these different roles comes from neurology, and from neuroelectrophysiology. MRI shows us the regions, but MRI doesn't see the roles.
Q: So, what do those regions have to do with the stimuli or behavior?
A: Well, it varies. And MRI can't tell. Correlation ain't necessarily causation; correlational neuroimaging cannot address causality. MRI shows us the regions that are correlated with the behavior, but not the different roles those regions serve.