Bugs in the software of MRI scanners call into question 40,000 scientific studies
Anatoly Alizar, "Giktimes"
A thorough study of the methodology of conducting research using magnetic resonance imaging invalidates the results of an entire branch of science. For several decades, physicians have used AFNI, SPM and FSL statistical programs to analyze fMRI data. As it turned out, due to incorrect algorithms, these programs can return up to 70% of false positive results instead of the expected 5%.
Thus, approximately 40,000 scientific papers published in recent decades on the basis of fMRI data are called into question overnight. In addition, the new validity assessment can have a strong impact on the interpretation of neuroimaging results.
Functional magnetic resonance imaging has been used in medicine for more than 25 years, and it is simply surprising that until now the most commonly used statistical methods that are used in MRI software have not been confirmed on real data, write the authors of the study published on June 27, 2016 in the journal Proceedings of the National Academy of Sciences (Eklund et al., Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates).
Statistical methods are the basis for interpreting fMRI results. This method of neuroimaging allows you to determine the activation of a certain area of the brain during its normal functioning under the influence of various physical factors (for example, body movement) and in various pathological conditions.
In the process of neuroimaging, based on the results of fMRI scanning, the brain is divided into tiny areas (voxels) based on the results of high-resolution scanning. The software then scans the voxels and combines them into "active" clusters that correspond to the activation of a specific area of the brain.
fMRI image with yellow clusters in which "increased activity is noted"
The problem is that this statistical procedure is performed incorrectly and, as it turned out, does not meet the scientific requirements for the maximum error of statistical data. The authors of the scientific paper write (in the press release of the Linking University Software for fMRI yield erroneous results - VM): "Our results show that the main cause of incorrect clustering are spatial autocorrelation functions that do not correspond to the assumed Gaussian distribution."
For example, a bug in the 3dClustSim software (part of the AFNI package) has been present for 15 years and was fixed only in May 2015, during the preparation of this study, write the authors of a scientific paper from the University of Linkoping (Sweden) and the University of Warwick (UK).
During the verification of the results of the programs, the data of functional magnetic resonance imaging at rest of 499 healthy patients from the control group were compared to obtain 3 million tomograms. It turned out that the percentage of false positive results is much higher than the estimated 5% and reaches 70%. The researchers conclude that parametric statistical methods, which are used in popular software packages SPM, FSL and AFNI, are actually unsuitable for cluster analysis of fMRI results, since the results do not correspond to the normal distribution.
This is what happens when research is conducted without checking the correctness of medical instruments for compliance with scientific standards of statistics.
Authors of scientific papers from the medical field will probably now have to conduct research again in order to repeat the result obtained using reliable tools with valid statistical methods.
So when you meet another study with fMRI results, you should take these results with caution. There is even an XKCD comic on this topic.
Surprisingly, the comic was drawn long before the study of statistical methods for clustering fMRI voxels by scientists from the University of Linkoping and the University of Warwick.
Portal "Eternal youth" http://vechnayamolodost.ru