Neuroscientists commonly do brain scan studies that use small study group sizes of fewer than 20 subjects per study group. Not long ago a press release from the University of Minnesota Twin Cities announced results which indicate that such small-sample correlation-seeking brain imaging experiments are utterly unreliable. The headline of the press release is "Brain studies show thousands of participants are needed for accurate results."
In the announcement we read this:
"Scientists rely on brain-wide association studies to measure brain structure and function—using MRI brain scans—and link them to complex characteristics such as personality, behavior, cognition, neurological conditions and mental illness. New research published March 16, 2022 in Nature from the University of Minnesota and Washington University School of Medicine in St. Louis...shows that most published brain-wide association studies are performed with too few participants to yield reliable findings."
The abstract of the paper in the science journal Nature can be read here. The paper is entitled, "Reproducible brain-wide association studies require thousands of individuals."
The press release tells us this:
"The study used publicly available data sets—involving a total of nearly 50,000 participants—to analyze a range of sample sizes and found:
- Brain-wide association studies need thousands of individuals to achieve higher reproducibility. Typical brain-wide association studies enroll just a few dozen people.
- So-called 'underpowered' studies are susceptible to uncovering strong but misleading associations by chance while missing real but weaker associations.
- Routinely underpowered brain-wide association studies result in a surplus of strong yet irreproducible findings."
The paper was released in March, 2022, but there is so far no sign that universities and neuroscientists are paying much attention to its very important findings. Universities continue to release shoddy press releases making dubious claims about the results of low-quality neuroscience experiments that use MRI scanning, poorly designed experiments that use fewer than 15 subjects per study group. A recent example is a Carnegie Mellon University press release trying to insinuate that neuroscientists have found something about "how abstract concepts are represented in the brain." No such thing has happened, because the study was yet another brain scan study using way too small study group sizes.
A look at the paper reveals that the study group sizes were only 10 subjects per study group. Failing to follow any blinding protocol, failing to do any sample size calculation, and failing to report any effect size, the study offers no robust evidence for anything. For the sake of this poorly designed study, twenty subjects (10 in each study group) had their brains scanned for 1 hour with a 3T scanner that may have health risks that I discuss below. Besides wasting federal funds, such studies may actually be putting subjects at risk by exposing them to long unnecessary brain scans that may have negative health effects, particularly decades down the road.
Another poorly designed brain scan study is a study announced with the press release below. The red circling shows one of countless examples I can give of how university press releases shamelessly hype, distort and exaggerate (a far-fetched speculation that brains might possibly use quantum computation is announced as simply a finding that brains do use quantum computation):
The study in question needlessly subjected 40 subjects to brain scanning with a 3T scanner (possibly dangerous for reasons discussed below). It seems nothing worthwhile was accomplished in the way of good experimental science because of the poor design of the study. It was not a pre-registered study; the study did not follow any blinding protocol; and there was no sample size calculation to determine whether the study group sizes used were adequate. Failing to meet any of the main hallmarks of a good experimental study, the study fails to report any effect size, fails to report any statistical significance, and mentions study group sizes way too small, including one group of only seven subjects and another group of only five subjects. The study refers to "1000 scans," causing me to worry about how much radiation these subjects were needlessly subjected to for a study so poorly designed.
An example of a poorly designed study which may have needlessly put many at risk was the study "Human cerebellum and corticocerebellar connections involved in emotional memory enhancement." Unlike the vast majority of neuroscience experiments, this study used a large sample size, and consisted of 1418 people who were brain-scanned. Unfortunately, the results are pretty worthless, because of a failure of the scientists to follow best practices. The paper makes no mention of a blinding protocol, an essential for a paper like this to be taken seriously. The study was not a pre-registered study. There was no declaration before gathering of a specific hypothesis to be tested, and a protocol of how data would be gathered and analyzed. So the scientists were free to keep slicing and dicing the data until they squeezed out a little "statistical significance" from some nook or cranny. The authors failed to report any specific percent signal change produced during memory activity. They fail to report results of any impressive statistical significance. Their best result is a mere "p < 0.05," which is scarcely worth even reporting. Many (such as the seventy experts who authored this paper) think that the criteria for statistical significance should be tightened, so that nothing should be reported as being statistically significant unless it has p < 0.005.
The subjects undergoing a half hour of brain scanning were paid a mere 25 Swiss Francs per hour that they were scanned, a trifling sum to be paid for being exposed to a significant risk. The subjects were brain scanned for 30 minutes (a medical MRI scan takes maybe 15 minutes). The more dangerous type of 3T scanner was used, a scanner type that has the possible health risks described below.
We can imagine the people who showed up for such a poor payment, amounting effectively to maybe $5 or $10 an hour (when you take into account transportation time, the time needed to exclude subjects facing higher risks by scanning, the training time required before scanning, and the time waiting for an MRI scanner to become ready). In general only the financially neediest people would have been induced by such a paltry payment. Similarly, the poorly designed GABA study described in this post subjected children to unnecessary brain scans by a powerful 3T scanner, and paid them a mere 25 pounds (nowadays worth about 28 dollars). What parent would allow such a thing for so a small a sum? Perhaps only one so needy as to have trouble feeding his child adequately.
It is a dogma among neuroscientists that MRI scans are safe. But we should remember that neuroscientists are very dogmatic creatures who often repeat claims that are dubious and unproven (as you can tell by reading the posts on this blog). Do we really know that MRI scans are free of any risk?
One danger of MRI scans is well-known: the risk of the very strong magnets used by such machines causing some metal object to be hurled at a high speed, causing injury or death. In 2001 a six-year-old boy was killed in the US during an MRI scan, when the machine turned an oxygen canister into a flying projectile. There is always a risk of lingering psychological trauma when certain people are put in some noisy high-tech machine and told they must be silent and not move for a long time such as an hour. There is also the risk that the more powerful MRI scans may raise the risk of cancer in the person getting the scan.
In the wikipedia.org article for Functional Magnetic Resonance Imaging, we read the troubling passage below:
"Genotoxic (i.e., potentially carcinogenic) effects of MRI scanning have been demonstrated in vivo and in vitro, leading a recent review to recommend 'a need for further studies and prudent use in order to avoid unnecessary examinations, according to the precautionary principle'. In a comparison of genotoxic effects of MRI compared with those of CT scans, Knuuti et al. reported that even though the DNA damage detected after MRI was at a level comparable to that produced by scans using ionizing radiation (low-dose coronary CT angiography, nuclear imaging, and X-ray angiography), differences in the mechanism by which this damage takes place suggests that the cancer risk of MRI, if any, is unknown."
The 2009 study here ("Genotoxic effects of 3 T
magnetic resonance imaging in cultured human lymphocytes")
cautions about the use of a high-intensity
("3T and above") MRI, and states that
"potential health risks are implied in the MRI and especially
HF MRI environment due to high-static
magnetic fields, fast gradient magnetic fields, and strong
radiofrequency electromagnetic fields," also noting that
"these results suggest that exposure to 3 T MRI induces
genotoxic effects in human lymphocytes," referring to effects
that may cause cancer.
A paper tells us the following about the newer twice-as-powerful
3T MRI machines that have been replacing the older 1.5T MRI
machines, suggesting their magnetic fields are much stronger than
the strength needed to lift a car:
"The main magnetic field of a 3T system is 60,000 times
the earth's magnet field. The strength of electromagnets
used to pick up cars in junk yards is about the field strength
of MRI systems with field strengths from 1.5-2.0T.
It is strong enough to pull fork-lift tires off of machinery,
pull heavy-duty floor buffers and mop buckets into
the bore of the magnet, pull stretchers across the room
and turn oxygen bottles into flying projectiles reaching
speeds in excess of 40 miles per hour."
A 2021 paper on MRI safety makes the not-very-reassuring claim that "no conclusive proof of harmful biological effects has been found to be caused by the static magnetic field up to 7T." This sounds like what cigarette manufacturers told us for years between 1950 and 1970, that there was no conclusive proof that cigarettes cause cancer (now such conclusive proof exists). When there is no evidence at all that something is harmful, a person will say something like "there is not a shred of evidence that it is harmful." When there exists some evidence suggesting a danger, a person may claim that there is "no conclusive proof" of harm.
Needlessly putting the needy at risk, usually for the sake of junk science
Will some of these subjects who participated in the usual poorly-designed brain scan studies end up with cancer decades from now because they were subjected to 30 to 60 minutes of unneeded 3T MRI scanning which "induces genotoxic effects" according to the previously cited paper? We'll probably never know, because neuroscientists don't seem to keep track of the long-term health results of the people they have brain-scanned in their experiments. It's kind of a policy of "scan 'em and forget 'em." Our neuroscientists are fond of saying there is "no proof" that MRI imaging can be harmful, but that's because they are not doing the long-term patient health followup tracking to determine whether MRI imaging produces a greater risk of cancer over 30 years or 40 years.
Don't put me down as being anti-MRI (I've had an MRI myself, after being advised by a doctor to do so). In countless medical treatment cases, the benefits of an MRI scan are greater than the small risks. But people should not be put at risk by getting unnecessary brain scans solely for the sake of poorly designed studies that fail to prove anything because they followed Questionable Research Practices.
I am not at all suggesting anyone should avoid an MRI scan when a doctor recommends such a thing as medically advisable. But it is rather clear that in their zeal to load up their resumes with more and more brain scanning studies, our neuroscientists are rounding up too many paid subjects for unnecessary and potentially harmful brain scans. What is really tragic is that such a large fraction of experimental brain scan studies follow Questionable Research Practices so badly that they qualify as "junk science studies" failing to provide any robust evidence for anything important. It seems that very often human research subjects may be needlessly put at increased risk of cancer and other health dangers by being brain-scanned in scanners such as 3T MRIs, merely so that neuroscientists can round up more subjects for badly designed studies that do nothing to advance science because they fall very short of meeting the standards of good experimental science.
When neuroscientists say brain scans are safe, they are referring to how much health trouble is now observed in people whose brains are scanned. No one has done some 25-year longitudinal study on the topic of whether people whose brains were scanned with 3T MRIs have a higher chance of cancer 25 or 30 years in the future. 3T MRIs were only approved by the FDA in the year 2000.
A scientific paper states this, referring to 3T MRIs:
"An insufficient number of validated studies have been carried out to demonstrate the safety of high strength static magnetic field exposure (Shellock, 2009). While MRI has been used for many years in the clinic, at higher Tesla levels (over 3 Tesla) the technology is relatively novel. Even less information about potential negative health effects exists for specific populations such as pregnant women and children."
The 2022 paper here discusses a large range of health concerns involving MRI scanners, such as these:
"The strong static magnetic field (B0) of MRI scanners can attract and accelerate ferromagnetic objects toward the center of the machine and turn them into dangerous projectiles...The radiofrequency (RF) field that is created by RF-coils can potentially cause tissue heating, especially in the presence of implants....The Time-varying fast-switching gradient magnetic field function is a spatial encoding of the MRI signal and can stimulate muscles or peripheral nerves and induce implant heating. They also produce noise in the MRI scanner space, which can reach levels of 100 dB or more and damage the hearing system."
If I were an ethical advisor asked to approve proposals for brain experiments, I would have the following rules:
- I would never approve the use of human brain scanning for any experimental study that used fewer than 25 subjects for any of its study groups, because such studies are way too likely to produce false alarms. I would never approve the use of any brain scanning experiment that did not include a sample size calculation to determine an adequate sample size that was used as a minimum for each study group.
- I would never approve the use of human brain scanning for any experimental study that had not published publicly a detailed research plan, including a precise hypothesis to be tested, along with a very exact and detailed description of how data would be gathered and analyzed. We should not be putting people at risk for studies that do not follow best practices.
- I would never approve the use of human brain scanning for any experimental study that had not published publicly a detailed blinding protocol to be followed, discussing exactly how blinding techniques would be used to reduce the risk of experimenter bias in which the experimenter "sees what he wants to see." We should not be putting people at risk for studies that do not follow best practices.
- I would insist that any consent form signed by a subject to be brain scanned would include a detailed discussion of the reasons why brain scanning might be potentially hazardous, with negative effects appearing far in the future, along with a fair discussion of the scientific literature suggesting such hazards. Currently a large fraction of such consent forms fail to frankly discuss such risk.
- I would never approve the use of any brain scanning on children in an experiment that did not absolutely require the participation of children.
I strongly advise all parents never to let their children participate in any brain scanning experimental study unless a doctor has told them that the brain scan is medically advisable solely for the health of the child. I advise adults not to participate in any brain scanning experimental study unless they have read something that gives them warrant for believing that the experimenters are following best experimental practices, and that there will be not be a very high chance that the adults will be undergoing unnecessary health risks for the sake of some "bad practices" poorly designed "fishing expedition" experiment that does not advance human understanding. If a neuroscientist looking for research subjects tells you that brain scans are perfectly safe, remember that many neuroscientists often dogmatically make claims that are unproven or doubtful, and often pretend to know things they do not actually know (see this site for very many examples).
I also strongly advise anyone who participated in any brain scanning experiment to permanently keep very careful records of their participation, to find out and write down the name of the scientific paper corresponding to the study, to write down and keep the names of any scientists or helpers they were involved with, to permanently keep a copy of any forms they signed, and to keep a careful log of any health problems they have. Such information may be useful should such a person decide to file a lawsuit.
When we examine the history of MRI scans, we see a history of overconfidence, and authorities dogmatically asserting that "MRI scans are perfectly safe," when they did not actually know whether they were perfectly safe. Not many years ago there arose the great "contrast agent" scandal. Scientists began to learn that what are called "contrast agent" MRI scans (given to 30 million people annually) may not be so safe. In such "contrast agent" scans, a subject is given an injection that increases the visual contrast of the MRI scan. For a long time, the main substance in such an injection was gadolinium. A mainstream cancer web site states, "Tissue and autopsy reports have also confirmed that gadolinium can accumulate in the brain and other organs." The results can be a health disaster, as described here. A 2019 Science Daily story says, "New contrast agent could make MRIs safer," letting us know that many of them previously were not so safe. On the same Science Daily web site, we read a 2017 news story with the title "MRI contrast agents accumulate in the brain." A 2020 paper ("Side Effect of Gadolinium MRI Contrast Agents") says this:
"Until recently, it was believed that gadolinium is effectively cleared within 24 hours after intravenous injection, and that it does not have any harmful effects on the human body. However, recent studies on animals and analyses of clinical data have indicated that gadolinium is retained in the body for many years post-administration, and may cause various diseases."
Neuroscientists extensively used such contrast agents (as described here), very often putting human subjects at risk for the sake of junk poorly designed studies falling far short of the best experimental practices. All the while, many of our experts were making the untrue claim that "MRI scans are perfectly safe," a statement which was not clearly true for the large fraction of MRI studies that used gadolinium contrast agents. You can do a Google search for "gadolinium deposition" to learn more about this issue.
A while ago there was published a scientific paper entitled "The effects of repeated MRI on chromosomal damage." Despite making in its abstract the claim that "MRI is a safe imaging technique," the paper finds results that are worrying. We are told, "The total number of damaged cells increased by 3.2% (95% CI 1.5–4.8%) per MRI (Fig. 2d–h) (p< 0.001); this increase being higher during the first ten MRI sessions than during the last ten ones."
Is there any protocol in place to prevent subjects from being used more than once in a neuroscience brain scan study? Apparently not, because Table 1 of the paper above mentions some subjects of the Human Brain Project being given more than 25 MRI brain scans, none of them medically necessary. The blase attitude of experimental neuroscientists towards health risks to their brain scan subjects is very appalling. Not only do they fail to track the long-term health of the subjects scanned ("scan 'em and forget 'em,") but also seem to pay no attention to how many times their subjects have been scanned, ignoring cumulative risk.
Most appallingly, our neuroscientists seem to fail to honestly notify their human subjects of the risks they are undergoing before engaging in possibly dangerous 3T brain scans so that the financially needy subjects can earn trifling sums. After doing a Google search for "brain scan experiment consent form," I find several actual consent forms and "model" consent forms containing dishonest language, such as the claim that there are "no known significant risks or side effect associated with MRI scans." That is not honest language, given the statements I have discussed above. One "template" for brain scan studies recommended by a university has these ridiculously inconsistent statements (the first untrue statement being contradicted by the rest of the statements):
"There are no known significant risks or side effects associated with MRI scans...There is a risk if metal objects are near the MRI because they can be drawn into the MRI scanner
and that could hurt someone in or near the machine...There may be risks associated with this study that we do not know about. In spite of all the care and precautions taken by the investigators, you might develop medical complications from participating in this study."
One horrid form by a major university asks participants to be scanned for an hour in an MRI machine as a "dry run" to test the machine or its settings. The form states this:
"The procedure may involve unexpected
risks that are impossible to predict. These unforeseen risks may
affect you during your participation in the procedure and/or at some
point in the future...You
will not be helped by participating in this procedure....You will not be paid for participating
in this study....If you are hurt as a result
of participating in the 'dry run exam', we have no plans to pay
you for lost wages, disability, or discomfort."
In general, the MRI consent forms I looked at totally failed to warn participants of the risk of cell damage and the increased chance of getting cancer as a result of a 1 hour 3T brain scan, something everyone should be warned of. As mentioned above, the 2022 paper "The effects of repeated brain MRI on chromosomal damage" found that "The total number of damaged cells increased by 3.2% (95% CI 1.5–4.8%) per MRI." The paper was referring to "DNA breaks" that have a possibility of increasing cancer risk. The paper referred to 90-minute 3T scans much longer than the average diagnostic MRI brain scan, which takes maybe 15 minutes and presumably damages much less than 1% of cells.
A paper entitled "A massive 7T fMRI dataset to bridge
3 cognitive neuroscience and artificial intelligence" discusses some data collection in which eight subjects were brain scanned 30 to 40 times with 7T scanners twice as powerful as the 3T scanners mentioned above, with each scan being about an hour long. The paper states this:
"The total number of 7T fMRI scan sessions were 43, 43, 35, 33, 43, 35, 43, and 33 for subj01–subj08, respectively. The average number of hours of resting-state fMRI conducted for each subject was 2.0 hours, and the average number of hours of task-based fMRI conducted for each subject was 38.5 hours."
This was in addition to other 3T scans the subjects were given. The paper makes no mention of any consideration of health risks to these people, who received only $30 per hour for the medically unnecessary scans. A 7T scanner would presumably have more than twice the risks of the 3T scanners discussed above.
Postscript: The latest example of needless risk to subjects is a study with a preprint entitled "Semantic reconstruction of continuous language from non-invasive brain recordings." The study failed to show any good evidence for anything important, as it used a way too-small study group size of only seven subjects (15 subjects per study group is the minimum for a moderately impressive result). Following Questionable Research Practices, the scientists report no sample size calculation, no blinding protocol, no pre-registration, no control group, and no effect size. The only "statistical significance" reported is what smells like "p-hacking" kind of results of the bare minimum for publication (merely p < .05). For these basically worthless results, seven subjects endured something like 16 hours of brain scanning with a 3T scanner, which is more than 30 times longer than they would have had for a diagnostic MRI. Senselessly, this study has been reported by our ever-credulous science press as some case of reading thoughts by brain scanning. It is no evidence of any such thing.
A 2005 article in Nature discusses second brains scans required in NIH-funded experiments, apparently to help clarify which subjects have brain anomalies that need to be reported to physicians. We read this about some workshop of "about 50 scientists, physicians, lawyers and ethicists":
"The NIH goes one step further by requiring its on-campus investigators to perform a clinical scan of every research subject, in addition to any research scans. But workshop participants agreed that this was not a good idea because it can expose people to unnecessary risks from extra procedures."
So the scientists seemed to agree that having two brain scans was exposing subjects to "unnecessary risks." So why are we having so many poorly designed neuroscience experiments in which so many subjects are needlessly subjected to such "unnecessary risks"? And why are some subjects being subjected to more than ten brain scans in such poorly designed experiments?
No comments:
Post a Comment