Some parts of the brain “light up” when we feel certain feelings, or listen to music, or tackle math problems. Certainly you’ve stumbled upon such news, given how frequently they end up in mainstream media. The technique used for these studies (and many others in neurosciences) is called functional Magnetic Resonance Imaging (or fMRI), which is an amazing thing, but also seems to have a few issues. I think we’ll be hearing about it in the near future, so it’s worth knowing what it’s all about.
An MRI machine. CC-BY-NC Penn State, via Flickr.
Let’s start from the beginning. The magnetic resonance imaging—MRI, the stuff sportspeople get done to assess injuries—uses magnetism and resonance (you don’t say!), or the unusual reaction of an object or material to stimuli of a specific frequency.
If, for example, we push someone on a swing every time they arrive all the way back, we’ll make them go higher than if we just pushed at random times. Simplifying (a lot), the MRI uses radio waves to push hydrogen atoms, of which there’s plenty in tissues rich in fat and water, like the brain.
Their nuclei have spin, a property that makes them react to magnetic fields somewhat like a compass would. The MRI machine applies a strong magnetic field, which aligns all the spins, then it hits them hits the atoms with a pulse of radio waves. If its frequency is just right (called resonance frequency), the wave will flip a few spins (not their atoms!) opposite to the magnetic field.
As soon as the pulse stops, it’s all back to normal, and atoms that flipped release a little energy. Recording these emissions with an antenna it is possible to distinguish tissues with different water contents, for example, different parts and layers of the brain, and build an image.
Simplified sketch of an MRI. The atoms (red balls) align along the green magnetic field until the magenta wave flips some of them. As soon as they can, atoms fall back to their original state, emitting energy recorded by the blue antenna. Credit: howequipmentworks.com
fMRI works by rapidly taking a lot of MRI pictures. Analyzing them we can understand what parts of the brain were more active at various time, because oxygenated blood rushes to these areas, producin a slightly different signal from that of “used” blood being flushed out.
The idea is simply genius. However, some recent studies urge caution and intense scrutiny on the statistical analysis used to process the images. In one of these studies, for example, a dead salmon seemed to react to pictures of humans.
That does not mean that the technique is rubbish, it just means we need to be careful. These studies are fundamental for research, because they let us identify issues. Only this way can we know what we’re doing and that we’re making the fullest and rightest use of the amazing results of spectacular techniques like fMRI.