Bug in software of MRI (Nuclear Magnetic Resonance Imaging), invalidating all research results of the past year or so

ByRoger Mommaerts

Nuclear magnetic resonanceBy using the phenomenon it is possible to make information in the living body image "MRI(Nuclear magnetic resonance imaging)"is. Unlike CT scan, since radiation is not used, there is a safety advantage that there is no radiation exposure, but a bug has been found on the software adopted by this MRI, and about 40,000 MRI images are displayed It became clear that the research results used are invalid.

MRI software bugs could upend years of research • The Register
http://www.theregister.co.uk/2016/07/03/mri_software_bugs_could_upend_years_of_research/


Magnetic resonance imaging (MRI) is frequently used for imaging cross-sectional images of the human brain, and so far, many MRI images have been used in brain research. However, "Results of research using MRI images" in the past year will be invalidated entirely. The cause is extremely simple, as a bug was found on the software used in MRI. In a research that uses scientific evidence for high resolution scanning of the brain using MRI,VoxelThis software is scanned by disassembling it. It is clear that there was a bug on this software.

We have discovered bugs in Sweden and the UK scientists and pointed out that there are many possibilities that misdiagnosis will be done with software currently being used in MRI. Also,Proceedings of the National Academy of SciencesIn thisAn article that pointed out facts"In general software packages for fMRI analysis (SPM, FSL, AFNI), there is a possibility of false detection of up to 70%".

ByRoger Mommaerts

According to a bug found in this software for MRI, the validity of the research result using about 40,000 fMRIs will be suspected, which will have a big influence on judgment of neuroimaging test results . In addition, the bug was found in the software package called "3dClustSim", and it seems that the bug occurred in May 2015.