Orthopedic implants cause significant artifact in MRI. Here, we will make a classification of these artifacts. Then approaches to minimize the susceptibility related artifacts are desciribed, such as the use of "wide band" sequences, view angle tilting and multi-spectral imaging methods.
Goals:
Several hundreds of thousands people per year are implanted with metal orthopedic instrumentation every year, in addition to people receiving aneurism clips, vascular stents, and pacemakers, etc… Many orthopedic implants are considered MR conditional, and do not exclude patients from being examined using MRI. Here, we will focus on the technical aspects of imaging near metal implants and leave the safety aspects such as potential heating, mechanical and device malfunctioning for discussion elsewhere.
The effects of orthopedic instrumentation on MRI are related to the differences of the implant material and tissue in susceptibility and conductivity. The differences in magnetic volume susceptibility lead to local variations of the static magnetic field $$$\Delta B_0$$$, which disturb the image encoding process (1-3). Careful selection of image encoding method and parameters can minimize the effect of these field errors, such that a useful image is obtained even in relative proximity of the implant. The conductivity of the implant causes interactions with the time varying electromagnetic fields, i.e. the RF-fields and the gradient fields (4-6).
Main field inhomogeneity: the local field gradient description.
For RF-refocused sequences useful observations can be made from describing the main magnetic field inhomogeneity $$$\Delta B_0$$$ as a field offset with local field gradients, such that:
$$ \Delta B_0(r+\Delta r,p+\Delta p,s+\Delta s)=\Delta B_0(r,p,s)+G_r\Delta r+G_p\Delta p+G_s\Delta s$$
This implies that during slice excitation the slice thickness is scaled by a factor $$$ \lambda=\frac{G_{Slice}}{G_{Slice}+G_s}$$$, and that the selected slice is centered away from its intended location by $$$-\Delta B_0(r,p,s)/G_{Slice}$$$. In the readout direction, assuming $$$G_{Read}>0$$$, all excited signal is displaced from its in plane location by a distance $$$\Delta B_0(r,p,s)/G_{Read}$$$ which is visible as signal loss for $$$G_r>0$$$ and signal pile up when $$$G_r<0$$$ . When $$$G_{Read}+G_r<0$$$ the position in the readout direction even becomes ambiguous and signal from separate locations will map into the same voxel in the image. No geometrical distortion occurs in the phase encoded direction(s). From the equations it can be seen that geometrical distortions and slice thickness variation can be countered by increasing the slice select and readout gradients, which is sometimes called the “wide-band” approach for metal artifact reduction.
Main field inhomogeneity as a frequency distribution
While the local field gradient description gives us useful insight, it does not apply generally to the field disturbances surrounding an implant. Instead, we may look at the anatomy as a region containing signal with a range of frequency offsets that result in encoding errors during excitation and readout: signal is excited as long as it meets the excitation condition $$$\mid\Delta f_0(r,p,s)+\bar{\gamma}G_{Slice}s\mid<BW_{RF}/2$$$, and is then displaced in readout direction according to $$$r'=r+\frac{\Delta f_0(r,p,s)}{\bar{\gamma}G_{Read}} $$$.
View-angle-tilting cancels one with the other by applying the same gradient in slice selection direction during excitation and encoding (7). However, view-angle tilting does not correct for slice thickness variations. SEMAC (8) resolved the slice distortions by a second phase encoding in the slice direction. MAVRIC (9) can be seen as an attempt to avoid the volume selection altogether, and use a 3D imaging approach with two phase encoding directions that are undistorted. However, since RF-pulse bandwidths are smaller than the range of frequency offsets, the sequence had to be repeated at different center frequencies, the so-called spectral bins. Building on this, a volume selective variation of MAVRIC was proposed (10), or rather a SEMAC-MAVRIC hybrid that addressed the limitations of MAVRIC with respect to folding artifacts in through plane direction. These techniques were coined Multispectral imaging (MSI) techniques, and try to cover the 3 spatial dimensions plus the spectral (frequency-offset) dimension to encode signal in a geometrically correct fashion so as to minimize metal artifacts. Because the frequency offsets as well as locations in the z-dimensions are scanned to complete the image, it is useful to represent MSI coverage in a hybrid B0-z diagram that shows the frequency and z-coverage of the method.
The images, that have three spatial and one spectral dimension, will then have to combined into a single 3D image, merging the information from different frequency offsets that originate from the same voxel. Here, sum-of-squares and complex addition are straightforward options, but other options are being considered that may reduce bin-edge (ripple) artifacts, reduce signal offset in low signal areas, improve SNR or reduce blurring.
Besides a number of review and educational papers on how to reduce metal artefacts in a clinical setting, for the technically interested audience I listed useful references related to the course content. (1-12)
Papers on MR and susceptibility
1. Ludeke KM, Roschmann P, Tischler R. Susceptibility artefacts in NMR imaging. Magnetic resonance imaging. 1985;3(4):329-43. Epub 1985/01/01.
2. Reichenbach JR, Venkatesan R, Yablonskiy DA, Thompson MR, Lai S, Haacke EM. Theory and application of static field inhomogeneity effects in gradient-echo imaging. Journal of magnetic resonance imaging : JMRI. 1997;7(2):266-79. Epub 1997/03/01.
3. Schenck JF. The role of magnetic susceptibility in magnetic resonance imaging: MRI magnetic compatibility of the first and second kinds. Medical physics. 1996;23(6):815-50. Epub 1996/06/01.
Nonsusceptibility artifacts
4. Camacho CR, Plewes DB, Henkelman RM. Nonsusceptibility artifacts due to metallic objects in MR imaging. Journal of Magnetic Resonance Imaging. 1995;5(1):75-88.
5. Graf H, Lauer UA, Berger A, Schick F. RF artifacts caused by metallic implants or instruments which get more prominent at 3 T: an in vitro study. Magnetic resonance imaging. 2005;23(3):493-9. Epub 2005/05/03.
6. Graf H, Steidle G, Martirosian P, Lauer UA, Schick F. Metal artifacts caused by gradient switching. Magnetic resonance in medicine. 2005;54(1):231-4. Epub 2005/06/22.
Artifact reduction methods
7. Cho ZH, Kim DJ, Kim YK. Total inhomogeneity correction including chemical shifts and susceptibility by view angle tilting. Medical physics. 1988;15(1):7-11. Epub 1988/01/01.
8. Lu W, Pauly KB, Gold GE, Pauly JM, Hargreaves BA. SEMAC: Slice Encoding for Metal Artifact Correction in MRI. Magnetic resonance in medicine. 2009;62(1):66-76. Epub 2009/03/10.
9. Koch KM, Lorbiecki JE, Hinks RS, King KF. A multispectral three-dimensional acquisition technique for imaging near metal implants. Magnetic resonance in medicine. 2009;61(2):381-90. Epub 2009/01/24.
10. Koch KM, Brau AC, Chen W, Gold GE, Hargreaves BA, Koff M, et al. Imaging near metal with a MAVRIC-SEMAC hybrid. Magnetic resonance in medicine. 2011;65(1):71-82. Epub 2010/10/29.
Simulations
11. Salomir R, de Senneville BD, Moonen CTW. A fast calculation method for magnetic field inhomogeneity due to an arbitrary distribution of bulk susceptibility. Concepts in Magnetic Resonance Part B: Magnetic Resonance Engineering. 2003;19B(1):26-34.
12. Bakker CJ, Bhagwandien R, Moerland MA, Ramos LM. Simulation of susceptibility artifacts in 2D and 3D Fourier transform spin-echo and gradient-echo magnetic resonance imaging. Magnetic resonance imaging. 1994;12(5):767-74. Epub 1994/01/01.