This is a fog-free translation of the article “Brain Imaging: fMRI 2.0” from Nature magazine. Translated on request :D http://www.nature.com/news/brain-imaging-fmri-2-0-1.10365
Oxygen-rich blood has magnetic properties so they can use a giant magnet to track the flow of blood through the brain; this is called an fMRI. Where there is increased blood flow, this shows increased brain activity; decreased blood flow shows decreased activity. This is great because they used to have to inject radioactive stuff into you to track brain activity and now they don’t.
Because it measures blood flow in the brain and not what the brain is doing directly, it’s not perfect; it can’t tell what brain cells are actually doing. Also the fact that blood flow in the brain fluxuates naturally anyway can make interpreting results difficult and makes it easier to misunderstand or misuse the results.
They also don’t know whether blood flow shows what the brain is doing now, or what it’s preparing to do, or what it was doing, or some other things as yet undiscovered. It’s better to measure directly what brain cells are doing; so some people are investigating measuring the activity of each cell individually. They think they can do that because brain cells have electrical activity; and electrical activity produces magnetic activity. So with really sensetive magnets tuned just right, they might be able to measure it.
It’s hard to use fMRI pictures to tell what parts of the brain are associated with what and why, because you just get a general idea of “the front bit is doing something” – but you don’t know if it’s doing what you think it’s doing or reading the screen or imagining a holiday etc. Scientists want to find out how parts of the brain communicate with each other. Some folk are, instead of squashing all the data they get from an fMRI into one picture, looking at it bit by bit to see the pattern of activity. Looking at lots of patterns allows them to identify the brains response to certain things, so for example they know a certain pattern is a response to an image of a bird.
fMRIs pick up a lot of useless background noise that is meaningless, and this needs to be filtered from the meaningful data. Scientists are thus working on ways to get more meaningful data and less noise; some ways are by using stronger magnets or injecting a substance into the blood which reacts better to magnets. They are also working on the best ways to filter the meaningful data from the meaningless noise.
At the moment most fMRI data is an average of many, many scans of lots of people, which gives a better chance of spotting the pattern of what the brain is doing. Obviously it would be great if you could use an fMRI in a hospital setting to analyse an individual patient’s brain to see what is going on – not currently possible. Scientists are experimenting with collecting large samples of fMRI data on various people and using them as a comparison base, so they can compare one persons fMRI to this huge database and be able to tell what is going on. Right now, you can’t use an fMRI to analyse what is going on in one persons brain, because they can’t understand what it is saying without having lots of people to compare it to.