Pectus excavatum is often corrected by the Nuss procedure, in which a metal bar (usually made of stainless steel) is inserted horizontally into the chest cavity. MRI examination is important to assess the cardiac function with a focus on the right heart. The metal bar causes susceptibility artifacts in cine MRI. We explore the feasibility to improve the B0 field homogeneity using a field correction device to be placed on the surface of the chest utilizing permanent magnets. A device was constructed based on computer simulations. The device generates magnetic multipole fields and effectively improved B0 homogeneity in a phantom.
Materials and Methods
Typically, a Nuss bars is 300 mm long, 12.5 mm wide, 2.5 mm thick, bent to conform to the shape of the ribcage, and 20 mm beneath the chest surface. The total induced magnetic moment of the horizontal bar at 1.5T is 0.03 A∙m2, and is uniformly distributed along its length. The field correction device would be placed on the chest surface and provide multipole magnetic field 5 to mitigate the B0 inhomogeneity. In the area near the heart, permanent magnets are arranged in 7 parallel lines with magnetic moments alternately opposed to or aligned with the B0 field (see Figure 2). The spacing between the lines was 15 mm and the length was 180 mm. Further away from the heart, 3 lines (50 mm and 40 mm long, respectively) were used on right and left sides. Permanent magnets (NdFeB magnets, grade N30AH, Pacific PAC Technologies, Inc.) were custom made with different sizes and the magnetic moment was measured at 1.5T with the magnetization parallel and anti-parallel to the B0 field 6. Permanent magnets were mounted on a piece of rigid PVC plate which was bent to conform to the contour of the chest of a typical patient (Figure 3). Although the magnets were discrete entities on the device, they approximated lines of magnetic dipoles. Computer simulations were performed to determine the magnetic dipole moment line density mi of the correction magnets:
$$ min. \int_{ROI}^{ }(B_{0,NussBar}(x,y,z)-B_{0, Device}(x,y,z))^2dV-\lambda\cdot\sum_{z\geq0} m_i^2 $$
The first term minimized the field inhomogeneity over an ROI where uniform B0 is desired, and the second term is for regularization which required the total absolute value of magnetic moment of correction magnets to be small. Magnets below the center line (z<0) were excluded from the sum because they are not independent. The parameter λ balanced the two terms. The equation was readily solved by a least squares routine using IDL. The device was constructed assuming λ=7x10-15 T2mA-2 (medium λ in Figure 4), and was tested on a rectangular phantom using an FFE pulse sequence used for cine cardiac MRI on an Ingenia 1.5T scanner.
1. Oezcan S, Attenhofer Jost CH, Pfyffer M, et al. Pectus excavatum: echocardiography and cardiac MRI reveal frequent pericardial effusion and right-sided heart anomalies. Eur Heart J Cardiovasc Imaging. 2012;13(8):673-679.
2. Saleh RS, Finn JP, Fenchel M, et al. Cardiovascular magnetic resonance in patients with pectus excavatum compared with normal controls. J Cardiovasc Magn Reson. 2010;12:73.
3. Nuss D. Recent experiences with minimally invasive pectus excavatum repair "Nuss procedure". Jpn J Thorac Cardiovasc Surg. 2005;53(7):338-344.
4. Topper A, Polleichtner S, Zagrosek A, et al. Impact of surgical correction of pectus excavatum on cardiac function: insights on the right ventricle. A cardiovascular magnetic resonance studydagger. Interact Cardiovasc Thorac Surg. 2016;22(1):38-46.
5. Jackson JD. Classical Electrodynamics. 3rd ed: John Wiley & Sons, Inc.; 1999.
6. Li S, Wang ZJ, Leigh JS. Magnetic susceptibility measurement by MRI - a dipole field method. Paper presented at: ISMRM 1998; Sydney, Australia.
7. Wang ZJ, Park YJ, Seo Y, Morriss MC, Rollins NK. Reducing brain MRI artifacts caused by ferromagnetic orthodontic appliances using permanent magnets. Paper presented at: ISMRM 2016; Singapore.