It seems like every week a new study comes along touting the new ways scientists can "read your mind." How exactly do scientists and doctors manage to peer inside your skull? Find out in this handy guide to brain scans.

The science of neuroimaging (the fancy name for brain scans) is relatively new. We've been able to x-ray the brain and see the blood vessels running through it since the early 20th century, but modern brain scanning technology wasn't developed until the 1970s. Today's researchers and doctors have better tools, and more of them, when it comes to looking inside the brain.

Electroencephalogram (EEG) – This is one of the earliest brain scanning technologies, with research on humans dating back to the 1920s and 30s. The EEG scan relies on the fact that nerve impulses within the brain take the form of (or generate) tiny electric currents, which indirectly cause changes in electric current at the scalp. Electrodes placed on the scalp can detect these changes and create a readout that can be interpreted to discern brain activity.
EEGs have very fine time resolution, but essentially no spatial resolution. They don't indicate much about what's going on inside the brain other than a general awareness of shifting brain states. This scan still has important medical uses, however. Since we know what a "normal" brain's EEG readout looks like, a scan can help determine if a mental problem such as dementia or a coma is due to a physiological problem within the brain or some other issue. EEGs are also used in sleep studies, since they can show how a sleeping subject's brain patterns are being disrupted. They're crucial in diagnosing epilepsy, since different forms of epilepsy show very different EEG readings during seizures.

Computerized Axial Tomography (CAT) – The 1970s and 80s saw several major breakthroughs in brain scanning technology. CAT scans (usually called CT scans in modern usage) are based on x-ray imaging technology. Tomography – the ability to focus an x-ray on a single "slice" of the subject – was used to get better contrast on specific areas. The use of computers allowed multiple tomography scans to be combined into a representation of the entire body part.


A modern CT scan is like a spiral x-ray. As the patient slowly moves through the machine, x-ray guns and detectors turn around the patient's head (or other body part). Computers assemble the information into a detailed scan of the entire brain, providing an incredible view of the structures within the brain. CT scans are medically invaluable, since they let doctors see brain injuries, skull fractures, tumors, blood clots and brain bleeding. They are crucial in detecting strokes, and are even used when a brain biopsy is performed – the scan gives doctors a real-time look at where the needle is.

Positron Emission Tomography (PET) – PET scans pick up where CAT scans leave off, showing brain activity instead of brain structure. In fact, they are often performed together by the same machine, creating a composite view of the brain's internal condition. Instead of x-rays, PET scans detect radioactive glucose injected into the subject. Organs (including the brain) absorb glucose when they are active, and the radioactive glucose emits gamma rays during absorption. Detecting these rays allows doctors to see what parts of the brain are active. While a PET scan doesn't show brain structure, it does reveal localized brain activity, which in turn lets researchers see which parts of the brain are active during certain activities, or lets doctors see if certain parts of the brain are acting abnormally.

Magnetic Resonance Imaging (MRI) – An MRI scan doesn't use radioactive injections or x-rays. Instead, it uses a powerful magnetic field to line up all the hydrogen atoms in a person's body (which works because the atoms' nuclei are protons). Strong pulses of radio waves briefly change the protons' alignment before they fall back in line with magnetic field. As the protons shift alignment, they release small amounts of energy, which the scanner detects. The protons in different tissues realign at different speeds and emit different amounts of energy, which is the "resonance" part of MRI. During the course of the scan, a variety of signals trigger proton alignment changes, and a computer then builds the data into an image of the body part being scanned.


Functional Magnetic Resonance Imaging (fMRI) – fMRI does what regular MRI cannot do: it measures activity within the brain instead of structures. This is based on the principle that active brain areas have increased blood flow and use more oxygen than inactive areas. It turns out that blood has different magnetic properties depending on whether it's packed with oxygen or most of the oxygen in it has been used up. Thus, the fMRI scan can see which parts of the brain are sucking up all the oxygen, and are presumably the most active.

fMRI is seldom used in medical treatments – it's mostly restricted to research. This scan is primarily responsible for the recent boom in "when you think about sex, your lateral parietal cortex is active, therefore that's the sex part of the brain" type studies (note: I made that up, I have no idea what the lateral parietal cortex does). There are many critics of these studies, who suggest that they oversimplify a complex and interconnected system by making the brain seem like a modular toolbox of separate parts that suddenly spring into action when a particular function is required.

Magnetoencephalography (MEG) – This type of scan has the coolest sci-fi terminology, and it might be the brain scan of the future. It relies on the same basic principle as EEG: the brain generates tiny electrical signals when it's active. A MEG scan, however, doesn't read those potentials. Instead, it reads the slight magnetic fields caused by them. The problem is, those magnetic field changes are so small that they're completely overwhelmed by things like your computer monitor, the power lines running down your street or anyone nearby using a cell phone.

The solution is two-fold. First, shield the testing room from magnetic interference as much as possible. Second, create an incredibly sensitive magnetic field detector. That detector is known as a SQUID, which stands for superconducting quantum interference device. They have to be bathed in liquid helium to stay cool enough for superconductivity to occur, which makes them damn expensive. SQUIDs also show up in some MRI scanners, and in any number of science-fiction novels. During a MEG scan, the subject's head sits within a "helmet" of SQUIDs, which detects brain activity while the subject performs tasks or look at images. While the scan is better at seeing areas near the surface of the brain, and doesn't have the as much spatial resolution as an fMRI, it has incredible time resolution, allowing researchers to see brain activity almost in real-time.

In the next ten to twenty years, brain scanning technology will undoubtedly get better and cheaper. In 50 years, will we have the equivalent of true mind-reading machines? Of course it's hard to say, but consider this: in 1935, the computational power required for an MRI or CAT scan to operate successfully was completely unimaginable.

Top image by: National High Magnetic Field Laboratory