By Rylan Dray, Ph.D.
UNLIKE STANDARD Magnetic resonance imaging, which generates images of brain (and other bodily) structure for diagnostic purposes, fMRI (functional…) is employed for clinical purposes *and* research into mental processes using a method called BOLD (blood-oxygen-level-dependent) signaling. Neural activity in the brain is usually linked to increases in local cerebral blood flow, leading to an increases in oxygenation that generate the signal. That is, when an area of the brain is in use, blood flow to that region also increases, a phenomenon known as neurovascular coupling. Thus, by measuring changes in blood flow, brain activity is inferred. Clinical uses include assessment of the effects of stroke, trauma or degenerative disease on brain function, monitoring the growth and function of brain tumors, and guiding the planning of surgery and radiation therapy). In research, subjects generally perform a task in the machine to see which area of the brain lights up. In recent years, fMRI has provided new insights into the investigation of language, pain, learning, emotion, and the formation of memories.
The first evidence supporting a coupling between energy metabolism and blood flow in the brain was provided in the late 19th century by Sherrington & Roy. Although they showed that blood volume does change locally in the brain, it was still unclear whether the brain itself was responsible for mediating these changes. It took till 1948 to confirm that blood flow in the brain is regionally regulated by the brain itself. Kety & Schmidt demonstrated that when neurons use more oxygen, chemical signals cause nearby blood vessels to dilate. In 1990, the fMRI proper was developed when Kwong and Ogawa independently discovered and utilized the BOLD effect.
My favorite fMRI experiment is that of Harvard psychologist Joshua Greene who in 2001 published research showing that moral decision-making relies on emotion rather than on pure reason. His group used fMRI to scan subjects as they read about moral dilemmas, and found that brain areas associated with emotion lit up when the more vexing scenarios were considered. However, the use of neuroimaging in psychology has been criticized as being scientifically unsound (about which, more presently), some like neuroscientists Rodolpho Llinas and W. R. Uttal likening it to phrenology, the 19th century pseudoscience that claimed to divine personality traits from skull bumps corresponding to discrete, specialized areas of the underlying organs.
“We are peering into a black box and attempting to draw conclusions,” a neuropsychologist I worked with during my rehab counseling internship told me. “Tricky business at best.” He conceded, though, that Greene’s moral dilemmas study “has at least face validity.” In a personal communication, Joshua Greene said in response that while there is some merit to such criticism — and he agreed that there has been a fair amount of regrettable hype– this merely reflects the lack of development of the science. He expressed optimism that its present crudity will be overcome. “Part of what we’re doing now is just getting our bearings, developing a more detailed functional map of the brain. You can call all that ‘phrenology’, but if it’s accurate…then it’s valuable.”
A 2020 review study had bad news for scientists keen on using task-oriented fMRI to draw conclusions about any one person’s brain. Ahmed Hariri, a Duke University neuroscientist, had been using the technology for many years to predict a person’s patterns of thoughts and feelings during specific mental tasks but started to see unreliable results across individuals when they were tested more than once. His group then re-examined 56 published papers on fMRI data to gauge their reliability across 90 experiments. The team concluded that for 6 out of 7 measures of brain function, the correlation between tests taken about 4 months apart was poor. In their paper, they presented “convergent evidence demonstrating poor reliability of task-fMRI measures…Collectively, these findings demonstrate that common task-fMRI measures are not currently suitable for brain biomarker discovery of for individual-differences research.” They added that the situation could be remedied in part if tasks are developed “from the ground up to optimize reliable and valid measurement.”
Of course, test-retest reliability is hardly the only problem fMRI faces. In fact, one article has listed 26 “controversies and challenges” for the technology: Neurovascular coupling; draining veins; linearity; long duration; mental chronometry; negative signal changes; resting state source; dead fish activation; global signal regression; motion artifacts, and so on. Neurovascular coupling, as noted earlier, is the physiological basis of fMRI but the authors write, “We know that the [cerebral] flow response is robust and consistent. We know that in active areas, oxygenation in healthy brains always increases; however, we just don’t understand why it’s necessary.” Just one more comment about this: re: “dead fish activation” – In 2006, scientist Craig Bennett showed BOLD activation in a dead salmon’s brain. (He discovered this when whimsically calibrating the machine, substituting odd items for the standard balloon.) While it illustrated statistical weakness regarding the multiple comparisons problem, the authors concluded that “no matter what statistical test is used, the reality is that the signal and the noise are not fully understood; therefore, all are actually approximations of truth, subject to errors.”
In a 2014 paper titled, “Brain Images, Babies, and Bathwater,” cognitive neuroscientist Martha J. Farah reviewed some of the criticisms of functional resonance imaging, opining that “The concern that fMRI shows us blood oxygenation rather than neural activity directly should be weighed alongside the fact that little of what we call science involves direct observations of the subject matter of interest.” After illustrating her point with references to cosmologists, chemists, and climate scientists, she concludes, “Complaints that functional neuroimages do not ‘show’ brain activity appear to based on a naive view of science and its methods.”
Needless to say, this brings up several issues in the philosophy of science, but that will have to wait for another post.
~ Rylan Dray, Ph.D. – December 2022
neuropsychsoc 