|Home | About | Journals | Submit | Contact Us | Français|
To map the distribution of the magnetic force induced in the human head during magnetic resonance imaging (MRI) at 4 T for a large group of healthy volunteers.
The magnetic field distribution in the head of 100 men and 18 women was mapped using phase mapping techniques. Statistical parametric mapping methods using a family-wise error (FWE) corrected threshold P < 0.05 and region-of-interest analyses were used to assess the significance of the results.
Eyeballs, orbitofrontal and temporal cortices, subcallosal gyrus, anterior cingulate, midbrain, and brainstem (pons) are the brain regions most susceptible to magnetic force. The strength of the magnetic force density in the head was lower than 11.5 ± 5.3 N/m3 (right eyeball). The strength of the magnetic force density induced in occipital cortex varied linearly with the x-rotation (pitch) angle.
We found that the induced magnetic force is highly significant in the eyeballs, orbitofrontal and temporal cortices, subcallosal gyrus, anterior cingulate as well as midbrain and brainstem (pons), regardless of subjects' age or gender. The maximum induced magnetic force was 6 × 105 times weaker than the gravitational force; thus, biological effects of the magnetic force during imaging are not expected to be significant.
MAGNETIC RESONANCE IMAGING (MRI) has become one of the safest medical techniques (1). However, with the steady progress of MRI technology, from lower to higher magnetic field strengths, potential adverse effects of the increasing magnetic force on living tissues are calling for more attention (2). Because of safety concerns, there is a myriad of studies evaluating the magnetic force on medical devices exposed to MRI (reviewed in Refs. 2-4). However, less attention has been given to effects of the magnetic force on living tissues during MRI scanning procedures. While in an earlier study on positron emission tomography it was reported that no effects of a 4 T static magnetic field on brain glucose metabolism, which is a marker of brain function (5), a few studies have associated dizziness (3), vertigo (1), magnetophosphenes (4), and metallic taste (6) to exposure to high magnetic fields in humans, and the quantification of the spatial distribution of the magnetic force in humans and animals during MRI scans is still lacking. Our recent quantitative measurements of the magnetic force distribution in the head of one human subject during 4 T MRI (7) reveled that the magnetic force induced in tissues of the human head during MRI procedures was lower than 0.1% of the gravitational force applied on the same tissues, suggesting a minimal impact of the MRI static magnetic field on biological tissues (7). However, the induced magnetic force in tissues can vary due to brain shape differences across subjects and/or the orientation of the head with respect to the static magnetic field. Therefore, the purpose of the current study was to evaluate the magnetic force applied on biological tissues in the human head during standard MRI procedures in a large group of subjects (>100 healthy subjects). Based on our previous single-subject measurements (7) and previous numerical simulations by others (8), we hypothesized that the induced magnetic force during MRI scans would be highly significant across subjects in the eyeballs and in the vicinities of the sinus cavity (orbitofrontal cortex) and the temporal bones (temporal cortex). For this purpose we mapped the induced magnetic field and the induced magnetic force in the head of 118 healthy human subjects using 4 T MRI and phase-mapping techniques using a susceptibility-weighted MR pulse sequence (7,9,10).
A total of 118 healthy volunteers (100 males, 18 females; age = 35.5 ± 17.5), who provided written consent approved by the local Institutional Review Board, participated in this study. Subjects were included if they were older than 18 years and younger than 60 years and excluded for 1) history of head trauma or loss of consciousness (> 30 min); 2) abnormal vital signs at time of screening or history of major medical conditions; 3) pregnancy as confirmed with a urine test in all females; and 4) contraindications to MRI.
Imaging was carried out in a 4 T Varian/Siemens MRI scanner. The homogeneity of the static magnetic field was maximized for a spherical and uniform water phantom (ID = 15 cm) using a 1D-FID NMR experiment by means of first- and second-order active shimming channels; the magnetic field homogeneity was better than 0.05 ppm in the volume of the phantom. These shimming parameters were fixed across subjects to minimize shimming-related magnetic field gradient differences across subjects. A fast low angle shot (FLASH) pulse sequence (11) with 31 sagittal slices; 4 mm thickness; 1 mm gap; in-plane resolution = 1.56 mm; acquisition matrix = 128 × 128; TR = 0.3 seconds; TE1 = 4 msec, TE2 = 5 msec; total scan time 4 minutes; phase-encoding gradient step = 0.18 mT/m; slice selection gradient = 5.25 mT/m; readout gradient = 9.81 mT/m; acquisition bandwidth = 83.59 kHz was used to measure the magnetic force distribution in the subjects' heads. Anatomical images were acquired using a T1-weighted 3D-MDEFT sequence (12) (TE/TR = 7/15 msec, 0.94 × 0.94 × 1 mm spatial resolution, axial orientation, 256 readout and 192 × 96 phase-encoding steps, 16 min scan time), and a modified T2-weighted Hyperecho sequence (13) (TE/TR = 42/10,000 msec, echo train length = 16, 256 × 256 matrix size, 30 coronal slices, 0.86 × 0.86 mm in-plane resolution, 5 mm thickness, no gap, 2.5-minute scan time), which were carefully reviewed to rule out major morphological abnormalities in the brain.
The longitudinal component of the magnetic field distribution, Bz(r) = γ−1 ϕ(r, TE2 − TE1) * (TE2 − TE1)−1, was calculated from differential phase maps, ϕ(r, TE2 − TE1), resulting from the complex ratio between FLASH datasets with TE = 5 msec and TE = 4 msec. This allowed us to minimize phase wrapping artifacts. The magnetic field gradient maps were calculated from magnetic field differences between adjacent voxels:
The magnetic force acting on biological tissues, which is proportional to their isotropic magnetic susceptibility (14), χ 1, was calculated according to:
where V is the voxel volume, B0 is the static magnetic field strength, and μ0 stands for the vacuum permeability. A customized algorithm, computing Eqs.  and  voxel-by-voxel, was developed in IDL (ITT Visual Information Solutions, Boulder, CO) to calculate g and Fz.
The Statistical Parametric Mapping (SPM2; Wellcome Department of Cognitive Neurology, London, UK) package was used for imaging postprocessing. The magnitude image corresponding to TE = 4 msec was spatially normalized to the standard frame (Talairach) with a 12-parameters affine transformation (15) using the standard T2* template provided with the SPM2 package, and these parameters were used for the spatial normalization of the calculated maps (gx, gy, gz, and Fz) using a voxel size of 3 × 3 × 3 mm. An 8-mm full-width-half-maximum (FWHM) Gaussian kernel was used to smooth the maps of the induced magnetic field gradients and forces, which were subsequently entered into a general linear model in SPM2 for group analyses. Independent analyses were carried out for each gradient component (gi) and Fz. The statistical significance of the induced field gradients and forces was evaluated voxel-by-voxel across subjects using one-sample t-test (random-effect) models and a family-wise error (FWE) threshold Pcorr < 0.05, corrected for multiple comparisons with the continuous random field calculation implemented in SPM2.
The magnetic force distribution was further evaluated with ROI analyses to identify potential outliers that might influence linear regressions and to report average values in a volume comparable to the image smoothness (e.g., resolution elements, or “resels” (16) rather than single-voxel peak values. The volume of the resels was estimated using the random field calculation in SPM2 as a near isotropic volume with FWHM = 6 mm. Thus, the magnetic force distribution was evaluated across subjects in 19 ROIs using 9-mm isotropic cubic masks containing 27 imaging voxels (0.729 mL) and centered at the Talairach coordinates listed in Table 1. The strength and standard deviation of the magnetic field gradients and forces induced by tissue susceptibility within these fixed masks was calculated across subjects using a custom program written in IDL and used in subsequent correlation analyses with age and orientation of the head in the MRI scanner. A threshold Pcorr < 0.05, corrected for multiple comparisons (Bonferroni), was considered statistically significant in ROI analyses.
The magnitude image corresponding to TE = 4 msec was registered to a standard template using a rigid body transformation in SPM2. The calculated x-, y-, and z-rotation angles and the x-, y-, and z-translations for each subject were used in subsequent correlation analyses to evaluate head orientation-related effects on the distribution of the magnetic force in the head.
Fig. 1A illustrates the ROIs discussed in the text and listed in Table 1. The highly heterogeneous pattern of induced magnetic field gradients reflected susceptibility differences at air-tissue interfaces near the sinus cavity and the temporal bone (Fig. 1, Table 1). Across subjects, gz and gy had a symmetric distribution with respect to the middle sagittal plane; they were negative in the inferior temporal (ITG, BA 37), subcallosal (SG, BA 25), anterior cingulate (AC, BA 25), and medial frontal (MFG, BA 25, ventral AC) gyri (> − 1.15 mT/m) and positive in the eyeballs, the inferior frontal gyrus (IFG, BA 20, ventral temporal), the cerebellar tonsil, and MFG (BAs 10 and 11) (< 1.3 mT/m) (Fig. 1B, Table 1). On the other hand, gx exhibited an antisymmetric pattern (Fig. 1C, Table 1), being positive in the left ITG (< 0.3 mT/m) but negative in the right ITG (> − 0.53 mT/m).
The distribution of Fz was similar to that of gz as a consequence of Eq.  (Fig. 1E). The magnetic force was minimal in deep brain regions but maximal in the eyeballs (< 250 pN) and at the air/tissue interfaces near the sinus cavity and temporal bones (Fig. 1E). The SG (across individuals, Fz < 150 pN), MFG (< 80 pN), and AC (< 50 pN) were the brain regions exhibiting the highest magnetic force levels during MRI. The induced magnetic force in the human head in these regions was highly significant (T-score > 30; Pcorr < 0.0001; Fig. 1F). The average Fz values were not statistically different for men and women (Fig. 2) and there were no statistically significant correlations between age and Fz either.
The lingual gyrus (BA = 23) was the only brain region that showed a statistically significant linear correlation between Fz and the head-orientation parameters. Specifically, the induced magnetic force in the lingual gyrus increased with the x-rotation (pitch) angle (R = 0.4; Pcorr < 0.0001, Fig. 3, regression slope = 0.023 ± 0.005 Nm−3 deg−1).
There were no correlations between other head orientation parameters (x-, y-, and z-translations; y- and z-rotations) and Fz in the lingual gyrus, or between any of the six head orientation parameters and Fz in any of the remaining ROIs in the brain.
Here we summarize our brain mapping results of the induced magnetic field and forces acting on biological tissues in the head of 118 healthy human subjects exposed to the homogeneous static magnetic field of a 4 T MRI scanner. As far as our knowledge reaches, this study is the first to assess the distribution of magnetic field gradients and forces in a large cohort of human subjects, thus providing an important database toward better assessment of MR safety and potential physiological effects of static magnetic fields in humans. Our main finding is that the magnetic force induced by magnetic susceptibility differences at air/tissue interfaces in the head is highly significant across all subjects. As demonstrated by our correlation analysis, only in certain regions such as the right lingual gyrus was the distribution of magnetic force significantly affected by the orientation of the head, relative to that of the magnetic field.
The magnetic force applied on weakly magnetic/diamagnetic biological tissues immersed in a perfectly homogenous external magnetic field is proportional to the local field strength, the magnetic susceptibility of the tissues, χ, and the inhomogeneity of the local magnetic field (ie, the static magnetic field gradient, g), which reflects the geometry of tissue magnetic susceptibility differences, Δχ (7,14,17).
The average magnetic force in the human head was maximal at the eyeballs and the orbitofrontal and temporal cortices (Fig. 1, Table 1), consistent with previous calculations of the induced static magnetic field in the human head (8) and previous findings in a single individual (7). These regions include the hypothalamus, which controls body temperature as well as other autonomic functions, and the vestibular system that contributes to our balance and our sense of spatial orientation. Thus, our results are also consistent with the higher occurrence of tonic vestibular asymmetry, hyperreactive caloric responses, and spontaneous nystagmus documented after 30 minutes exposure to 9.4 T static field in a small group of MRI workers (18). Vestibular dizziness has been shown to deactivate the middle/medial superior temporal area (MT/MST) (19), and the same area was found to be associated with taste perceptions using different taste stimuli (20). Thus, our findings on induced magnetic force across all subjects in the ITG (Fig. 1E,F) suggest that feelings of dizziness, vertigo, and “metallic taste” after exposure to high-field MRI could be associated with neurostimulation induced by magnetic force in the ITG and the other mentioned regions. However, at current magnetic fields no such symptoms were observed/reported in this cohort of subjects.
Lingual gyrus was the only brain region that showed a statistically significant linear correlation between Fz and head orientation parameters (x-rotation, or pitch) (Fig. 3). This finding could be associated with the sensation of vertigo in the MRI scanner. The sensation of vertigo is frequently experienced when subjects move in and out of the MRI scanner, and could result from electrical currents induced in moving brain tissues exposed to magnetic fields (14). However, when the head motion is perpendicular to the orientation of the magnetic field it can induce sensations of vertigo in humans and this has been postulated to arise from magnetohydrodynamic forces giving rise to similar effects as those seen in travel sickness (21). Since the lingual gyrus was shown to activate in previous functional MRI studies on vertigo (1,22), our findings on orientation-dependent induction of magnetic forces in the lingual gyrus suggest that the magnetic force acting in these brain regions could originate sensations of vertigo in MRI (3,4,22), although these sensations would be much weaker comparing to that caused by motion in the static magnetic field (e.g., 1,3,4,6,22,23). Finally. This study does not support significant gender or aging differences in the magnetic force distribution in the human heads.
The amplitude of the imaging gradients and the acquisition bandwidth limited the accuracy of the measurements (estimated systematic error ≈ 3%). Furthermore, the measurements reflected average values within the imaging voxels. Thus, future studies using higher spatial resolution might allow detection of higher peak values for the magnetic force in the human head. This study evaluated the induced force during real MRI scan only; it does not measure the magnetic force while subjects entering or exiting the magnet, which at 4 T could be 300 times higher than that reported in this study depending on the motion speed of the body.
In conclusion, we assessed the magnetic force distribution in the head of 118 healthy humans exposed to 4 T magnetic field MRI using statistical parametric mapping techniques. We found that the induced magnetic force is highly significant in the eyeballs, orbito-frontal and temporal cortices, subcallosal gyrus, anterior cingulate as well as midbrain and brainstem (pons), regardless of subjects age or gender. The induced magnetic force was 6 × 105 times weaker than the Earth's gravitational force. While biological damage caused by the induced magnetic force in the brain during MRI scanning seems unlikely, at least under current strength of MRI technology, the potential effects of the magnetic force on brain function merits further evaluation since it could affect activation/deactivation signal observed with fMRI studies.
We thank two anonymous reviewers for many constructive comments and useful suggestions on the article.
This article was authored by Brookhaven Science Associates, LLC under Contract No. DE-AC02-98CHI-886 with the U.S. Department of Energy. The United States Government retains, and the publisher, by accepting the article for publication, acknowledges, a world-wide license to publish or reproduce the published form of this article, or allow others to do so, for United States Government purposes.