The functional magnetic resonance imaging (fMRI) is one of the imaging techniques most useful to study the brain activity.
In fact, you will have found that in most of our articles on neurosciences We tell you that thanks to it it was possible to see which parts of the brain intervene in actions such as sleep, hunger, love, fear Okay, and if we told you that all that was a lie?
Saying so is a bit of a stretch, but I do know that according to a study recently published in PNAS, a total of 40,000 studies carried out during 15 years could have resulted in a falsification of results due to errors in the magnetic resonance software.
What is functional magnetic resonance for?
The functional magnetic resonance It is a very useful tool, since it serves to know through images the brain regions who are doing a concrete task in a certain moment.
The entire procedure takes place on the same device used in the diagnostic resonancesbut they are done small changes in hardware and software that will allow us to analyze brain activity.
For this, the patient is first made to perform the homework whose brain activity you want to study (talk, raise an arm, look at a specific photo). This will lead to a vasodilation of the brain region involved, due to the need for a increased glucose supply, so that in this zone the concentration of a substance, called deoxyhemoglobinwhich in turn change the local magnetism, what to be detected by the resonator.
Finally the obtained results are analyzed by a software concrete that lead to statistical data that they finally are collated by researchers.
How did they check if the MRI software falsified the results?
Magnetic resonances are generally analyzed with one of the three most important software, called SMP, FSL and AFNISo it was precisely on them that these scientists from Linkping University in Sweden focused.
To do this, they took resting data of 499 patients from around the world and compared the results of their magnetic resonances so they got 3 million random comparisons.
What is expected in case the procedure is being done well will be at most a 5% of what are known as false positivesbut the researchers found an astonishing 70% falsification of the data.
The results could not be considered reliable with a error rate so high, therefore the failures that gave rise to this imbalance were analyzed and corrected in 2015, when the article began to be written. However, by then they had been doing it for 15 years studies with that error, so it could be that a total of 40,000 investigations they had incorrect results.
In addition, until recently this problem was linked to the slowness of these computer programs and the high cost of resonances, which forced researchers to carry out studies with few patients and little margin to the repeat experiments.
Luckily, much progress has been made in this regard, what was previously processed in 10 years can now be done in 20 days and the results of the researchers are available online for the study of other scientists, who may afford not to perform some experiments that have already been carried out by others.
Therefore, this does not mean that everything we know so far about the brain is a lie, but it is true that some have been obtained incorrect results Due to something as simple as machine error, machines are never supposed to fail, but after all there is always a human behind them. And to err is part of us. The important thing is to know how to solve it in time.
