If you’ve been reading this blog (What? You haven’t? Shame on you!), then you know I often refer to fMRI (functional magnetic resonance imaging) experiments. The fMRI had its genesis at Mass General in 1991- when giant magnetics afforded us the ability to visualize activity in the brain. (This final result relied on the finding by Seiji Ogawa at Bell Labs that determined blood-oxygen-level dependence, the MRI contrast of blood deoxyhemoglobin [oxygen depleted red blood cells].)
We love using fMRI because it’s basically non-invasive, there’s no need to inject radioactive tracers into the bloodstream, and can provide extremely fine resolution. And, we take the fMRI results as gospel. But, are they?
You see, fMRI does not really measure brain activity- but brain blood flow (called the hemodynamic response). Now, don’t jump to the wrong conclusion- these phenomena are. indeed, related. Our neurons transmit their impulses via neurotransmitters and electricity; and blood flow must increase to the neurons to afford them this capability. (You do know that the brain comprises some 2% of our body mass- but demands some 15% of the cardiac output….)
And, until recently, we had no clue whether the “bright lights” of fMRI resulted from the input to or output from neurons- or something else entirely. Now, Drs. Logothetis, Pauls, MaAugath, Trinath, and Oeltermann of the Max Planck Institute answered part of the puzzle. In their Nature article, they monitored the fMRI signal simultaneously to measurements of the neuronal activity in monkey’s visual cortex. They found the fMRI signal was most related to the input to neurons (processing inputs is a major energy consuming function, so that would need more blood flow).
But, when we want to examine complex phenomena- like love- then we are relying on those algorithms that devise the connections between neurons and blood flow. And, there’s plenty of noise that is filtered out by the computer analyses. This noise is related to physiological fluctuations, bulk head motions, white noise (random actions), and system instabilities.
The problems occur when this noise is the bulk of the activity occurring, in that it is a result of physiological fluctuations, and not direct input processing by the neurons. (Hemodynamic response is relatively slow compared to neuronal responses.)
And, what happens when part of our brain is active- and that activity can denote either love or intense pain? (Do you really love your iPhone or do you hate hearing it ring?) It’s why these complex phenomena will not be fully comprehended until we find a more direct measurement system.
