MRI Safety Considerations for Permanent Magnet Implants in Muscle

Abstract

Background: Permanent magnet implants are used with several medical and assistive devices, such as cochlear implants, dental attachments, and prosthetic control, but raise caution for MR imaging. Previous work has evaluated several magnet implants for position and magnetization stability, as well as for image artifacts under MRI. Yet, the intramuscular magnets used for prosthetic control still require evaluation for potential MRI conditionality.

Purpose: To investigate the position and magnetization stability of and image artifacts from 3-mm-diameter spherical permanent magnets (B _r = 1.393 T, H _ci = 1.637 MA/m) implanted within muscle.

Study type: Prospective longitudinal study.

Animal model: Porcine; one animal, eight muscles.

Field strength/sequence: 0.55-T, 1.5-T/SE, GRE.

Assessment: Permanent magnets and nonmagnetic controls were implanted into eight muscles and exposed to 1.5-T MRI 36 days post-implantation. All sites were examined histologically for evidence of implant migration (acute fibrotic response or fibrotic capsule disruption). Benchtop studies evaluated worst-case demagnetization and image artifacts (artifact radius minus implant radius). The primary measure of position stability was histological examination interpreting characteristics of progressive skeletal muscle healing. Secondary position stability analysis was performed via CT imaging.

Statistical tests: Unpaired one-sided sign test with a significance level of 0.05. Demagnetization and imaging artifacts were summarized as maximums.

Results: Fibrotic capsules were similarly intact at permanent magnet and control sites (fibrotic capsule thicknesses: 20-550 μm [magnets], 20-220 μm [controls]). No effect of MRI exposure on implant migration was observed via secondary analysis (p = 0.965 [0.55-T], p = 0.996 [1.5-T]). Maximum demagnetization was 2.1% under 0.55-T exposure and 13.5% under 1.5-T exposure, and maximum image artifact was 71 mm at both imaging strengths.

Data conclusion: The permanent magnet implants used in this study were resistant to migration and substantial demagnetization under 0.55-T and 1.5-T MRI exposure and resulted in negligible image artifacts for critical organ imaging, suggesting that the presence of these implants does not preclude a patient from receiving MR imaging up to 1.5T.

Evidence level: N/A.

Technical efficacy: Stage 5: Improvements in patient care.

Keywords: MRI conditionality; intramuscular implants; magnetic implants; permanent magnet implants.

FIGURE 1

MR Imaging of a patient with permanent magnets implanted in a muscle. In the case of muscle interfaces such as magnetomicrometry, small permanent magnets are implanted into a patient's muscles for control over a prosthesis or other device. The effects of these magnets on the MRI image, and the effects of the MRI on the magnets, are considered in this work.

FIGURE 2

Analysis of permanent magnet and nonmagnetic control implant positions pre versus post MRI exposure. In each of eight muscles in one pig, we implanted a permanent magnet (highlighted in gold) and a nonmagnetic (tantalum) control (highlighted in blue). The main strategy we used to evaluate potential migration was histology. However, as a secondary evaluation of implant position stability, we also implanted three nonmagnetic (tantalum) position reference implants (highlighted in green) around each magnet. These reference implants provided a reference coordinate system to evaluate differences in position measurements within the computed tomography (CT) images for the magnets and controls. The left and right images represent transverse and dorsal views, respectively. The implants were placed in the left gluteus medius (GM), biceps femoris (BF), rectus femoris (RF), and semimembranosus (SM) muscles, and into the right SM, RF, BF, and GM muscles (listed here corresponding roughly to their clockwise arrangement in the left image).

FIGURE 3

Timeline for computed tomography (CT) imaging, MRI exposure, and histological analysis. Dashed lines indicate CT scans, and the dotted line indicates the time point for histological examination.

FIGURE 4

Magnetization strength of permanent magnets pre and post MRI exposure. (A) The magnetization strength (residual flux density) of each permanent magnet was measured using a custom Helmholtz coil measurement device. (B) Each magnet was then affixed at a predetermined angle in a custom fixture. (C) Each magnet was then exposed to MRI‐strength fields and gradients during imaging in the fixture at iso‐center (image shows fixture before movement into iso‐center), after which it was then measured again using the Helmholtz coil measurement device.

FIGURE 5

Histopathology Results. Spurr resin embedded cross sections of each implant site, sectioned at 5‐μm thickness and stained with hematoxylin and eosin (we sawed each sample in half through each implant after embedding, then ejected the implants before sectioning). The eight images on the left (panel A) are sections from the permanent magnet implant sites, and the eight images on the right (panel B) are sections from the nonmagnetic control implant sites. Each slide shows a 4.5 × 4.5 mm square of the implant cross section centered on the implant (a 1‐mm scale bar is provided in the bottom left hand corner of each slide). Rows correspond to muscles (semimembranosus, rectus femoris, biceps femoris, and gluteus medius) and columns correspond to the left versus right side of the animal. Note the fibrotic capsule around each implant site, and the remnants of the Parylene C coating, can be observed in many of the slides.

FIGURE 6

Measurement differences pre versus post MRI exposure. These cumulative distribution functions show the likelihood for the permanent magnet implants to have been measured with various perceived movements in comparison with the nonmagnetic control implants. The nonmagnetic control implants measurement changes were in general larger, likely due to the controls being located farther away from the reference implants, and thus more sensitive (through coordinate transformation) to changes in muscle morphology over time (the CT scans for the 0.55‐T exposure study were 8 days apart, while the CT scans for the 1.5‐T exposure study were only 3 days apart). For a table providing all data point values in these plots, see Table S1.

FIGURE 7

Demagnetization (empirical and simulated) magnetic spheres post exposure to 0.55‐T and 1.5‐T MRI. Remaining residual flux density post exposure to 0.55‐T and 1.5‐T MRI, with each magnet rigidly‐affixed at a given angle to the applied field.

FIGURE 8

Demagnetization (empirical) magnetic spheres under worst‐case exposure to 1.5T MRI at varying durations. Remaining residual flux density post exposure to 1.5T MRI for various durations, with the magnet rigidly‐affixed in direct opposition to the applied field. The horizontal axis is shown in minutes and is spaced logarithmically for ease of viewing, while the vertical axis shows the percentage of magnetization remaining post exposure of one of the fully‐magnetized permanent magnets to the 1.5‐T MRI.

FIGURE 9

Image artifacts resulting during 0.55‐T and 1.5‐T imaging. Image artifacts in 0.55‐T and 1.5‐T MRI images from the permanent magnet implant are shown, with 0.55‐T images on top, 1.5‐T images on bottom, spin echo images on the left, and gradient echo images on the right. Original images of the implant and of the background are shown in each plane, along with the thresholded difference image and the span markings (gold lines). The horizontal and vertical spans are listed next to each thresholded difference image. Note that the spans are labeled here, and that the image artifact extent is calculated as half the span minus the implant radius. All images show thresholded differences, thresholded at 30% change from pre to post introduction of the implant. Note that the image for each plane was collected separately, and also that the images in the figure are not scaled relative to one another. The axial, coronal, and sagittal images are shown from top to bottom in each of the panels.