Brain imaging studies are becoming increasingly prolific in the news and on websites, blogs and journals. Some of these have good internal and external validity, methodology and peer reviews, however a substantial amount do not. The study of brain imaging had its inception back in 1895 on the 8th November when Wilhelm Röntgen discovered X-rays, and since this time methods and analysis for investigating functional and anatomical aspects of the brain have ever since improved and are still doing so today. These days when a ‘brain scan’ is mentioned it is usually referring to Functional magnetic resonance imaging (fMRI) which claims to be a direct measure of brain activity. This is false. fMRI is indeed more informative than an X-ray but it is not a direct measure of brain activity as credible news sites like to claim. I shall refrain from name dropping here, but if you come across any such articles you will know. This article is to tell you what to look out for when reading about brain imaging studies, and will hopefully make an entertaining read too!
In reality fMRI is a measure of oxygenated blood; more precisely a measure of oxyhaemoglobin (oxygenated haemoglobin). The premise is that higher amounts of oxygen are required in the neurons that are doing the work (cognition requires certain neurons to perfuse so they need more oxygen). This neuronal work increases the concentration of oxyhaemoglobin at these sites and decreases deoxyhaemoglobin, and it is this that indicates localised activity in fMRI. As blood supply is limited, an average adult has 8 – 10 pints of blood in their system, so a decrease of oxyhaemoglobin indicates reduction of neuronal activity. All sounds good so far but, and there are some big ‘buts’ coming up.
fMRI signal does not distinguish between actual activity and inhibitory activity in the brain. This ambiguity exists because these types of responses generate increased levels of oxyhaemoglobin, and there is currently no way to tell them apart. So what might be described as a critical location of activation is actually the exact opposite. On top of this, when blood leaves the brain it tends to flow through the larger veins as opposed to smaller ones, this then affects the fMRI signal by indicating brain activity even though this is many neurons away from the actual location of activity.
If you have ever wondered as to why or how the brain uses approximately 20% of our energy resources, even when we don’t have deadlines to meet! This is because the brain demands a lot more energy (in the form of oxygenated blood including glucose) than what it actually consumes (so why do our brains get tired…?) In regards to brain scanning this means that what is being measured is our greedy brains demand for all this extra energy, rather than what we use during specific tasks which is what is measured in experiments.
“This Is Your Brain on Politics” (New York Times, Nov. 11) claimed that it is possible to read the minds of potential voters by watching their brain activity while they viewed the candidates. For a start neuroscience does not claim to read minds. Neuroscience is the study of the brain and the encompassing nervous system. It seeks to describe the functioning of the brain in the healthy and nonhealthy in order to cure and prevent diseases and disorders that lay in the brain.
One of my favourite articles by a neuroscientist critiquing fMRI is “fMRI Gets Slap in the Face with a Dead Fish”(Neuroskeptic blog). Back in 2009 neuroscientist Craig Bennett put a dead salmon in a fMRI scanner which was shown a series of photographs of humans in various social situations. Then just for effect he “asked” the salmon what emotion the human in each photograph was experiencing. Although this is high on the scale when it comes to comedy value, what is worrying is that the dead fish showed brain activation! Apparently dead salmon can distinguish between human emotions. This was due to the statistics used. Basically if multiple comparisons correction is not used then scientific methods manage to find neural activation in deceased fish. However if it is used then you don’t. What is more concerning is that many fMRI publications still report “uncorrected” results. Always look out for the corrected results if reading a paper!
So the bottom line is if reading a journal article make sure the data has been “corrected” and if reading else where then keep in mind that fMRI is not as direct as some like you to believe.