Modulatory effects of magnetic vestibular stimulation on resting-state networks can be explained by subject-specific orientation of inner-ear anatomy in the MR static magnetic field.

Strong static magnetic fields, as used in magnetic resonance imaging (MRI), stimulate the vestibular inner ear leading to a state of imbalance within the vestibular system that causes nystagmus. This magnetic vestibular stimulation (MVS) also modulates fluctuations of resting-state functional MRI (RS-fMRI) networks. MVS can be explained by a Lorentz force model, indicating that MVS is the result of the interaction of the static magnetic field strength and direction (called "B0 magnetic field" in MRI) with the inner ear's continuous endolymphatic ionic current. However, the high variability between subjects receiving MVS (measured as nystagmus slow-phase velocity and RS-fMRI amplitude modulations) despite matching head position, remains to be explained. Furthermore, within the imaging community, an "easy-to-acquire-and-use" proxy accounting for modulatory MVS effects in RS-fMRI fluctuations is needed. The present study uses MRI data of 60 healthy volunteers to examine the relationship between RS-fMRI fluctuations and the individual orientation of inner-ear anatomy within the static magnetic field of the MRI. The individual inner-ear anatomy and orientation were assessed via high-resolution anatomical CISS images and related to fluctuations of RS-fMRI networks previously associated with MVS. More specifically, we used a subject-specific proxy for MVS (pMVS) that corresponds to the orientation of the individual inner-ear anatomy within the static magnetic field direction (also called "z-direction" in MR imaging). We found that pMVS explained a considerable fraction of the total variance in RS-fMRI fluctuations (for instance, from 11% in the right cerebellum up to 36% in the cerebellar vermis). In addition to pMVS, we examined the angle of Reid's plane, as determined from anatomical imaging as an alternative and found that this angle (with the same sinus transformation as for pMVS) explained considerably less variance, e.g., from 2 to 16%. In our opinion, an excess variability due to MVS should generally be addressed in fMRI research analogous to nuisance regression for movement, pulsation, and respiration effects. We suggest using the pMVS parameter to deal with modulations of RS-fMRI fluctuations due to MVS. MVS-induced variance can easily be accounted by using high-resolution anatomical imaging of the inner ear and including the proposed pMVS parameter in fMRI group-level analysis.