[ EMF-Link Home
| Online Store
| Comments
| Up One Level ]
| TITLE: | Safety of Time Varying Magnetic Fields in MR Imaging. | ||
| Principal Investigator |
Om Gandhi, Sc.D., M.S.E. | University of Utah | |
| Health Relevance |
Other: Biophysics of Cell Field Interaction | ||
| Research Categories |
Biophysics of Cell Field Interaction | ||
| FY95 Funds | R01ES06292 $ 139,892 | Start Date 09/01/93 | End Date 08/31/96 |
| Rationale and Summary |
Over the past five to ten years, magnetic resonance imaging (MRI) has become an important tool for medical diagnostic applications. In the early days of MRI, most people thought that MRI systems posed no hazards to patients or operating personnel because no X-rays were involved in the imaging process and the magnetic field strengths used were not considered harmful. The frequencies were low enough that the radiofrequency (RF) magnetic fields did not induce appreciable internal current densities, and negligible RF heating was expected. Newer techniques are leading to the use of higher static magnetic fields; more rapidly switched gradient fields and higher RF magnetic fields. Although use of the increasingly stronger electromagnetic fields is causing concern about patient safety, to date, only simplified homogeneous spherical, cylindrical and disc models have been used to obtain induced current density distributions and rates of energy absorption (specific absorption rates or SARs). Since this knowledge is vitally important, particularly for the critical regions of the body, we will adapt the anatomically based modeling techniques that we have successfully used for electromagnetic dosimetry. We propose to use the previously tested highly efficient impedance method and its new generalization for SARs for RF magnetic fields, or realistic polarizations and variations typical of present and planned MRI systems. We also plan to use the newly developed modified finite-difference time-domain method, which has been used successfully for a wide variety of RF electromagnetic exposure conditions. For the various calculations we will use the new high-resolution model of a human volunteer based on MRI scans with resolutions of 1.875-3 mm) and its scaled versions where different scaling factors x, y, and z on the order of 0.6-1.0 will be taken to obtain models of different heights and weights representative of human adults and children. This latter step is needed since induced current densities and SARs are strongly dependent on the size and shape of the body. The current densities and SARs thus calculated will be compared with the safety guidelines suggested by USFDA and NRPB (U.K.). | ||
| Experimental Design and Exposure Conditions |
Calculation of SARs for Body Coil: For SAR calculations we have used a realistic geometry of a
16-run shielded bird cage body coil that is typically used in the MR imagers of today. Diameter and
length of the bird cage are 56 cm each, while the diameter of the shield is 65.5 cm. To obtain
circular polarization of the RF magnetic fields, each of the runs is assumed to be fed with equal
magnitude currents that are 22.5 out of phase relative to the previous rung. WE have used the
finite-difference time-domain method to calculate the induced electric and magnetic fields and
SARs first in -symmetric, glossy homogeneous cylinders to verify the accuracy of the computer
code, and subsequently in a 0.655-cm (nominal 1/4") resolution, 16-tissue, anatomically based
model of the human body.
Calculations of SARs for the Presently Used Head Coil: For these calculations we have taken the geometry and the parameters of the present-day commonly used head coil. Head coils are also made of a bird cage of 16-current-carrying rungs for which a progressive phase shift of 22.5 in the currents of the adjacent conductors is used to obtain circularly polarized RF magnetic fields. Diameter and length of the bird cage are 28.5 cm and 39.3 cm, respectively. Calculations of SARs for Proposed High-Frequency Head Coils: Tuned resonant-type higher- frequency head coils have recently been proposed and are being developed for improved NMR imaging with a higher signal-to-noise ratio and spatial resolution. We have used physical dimensions typical of such coils for frequencies up to the 400 MHZ that are planned for these coils. Currents Induced by Switched Gradient Magnetic Fields: We have used the impedance method to calculate the currents induced in a 0.655-cm (nominal 1/4") resolution, anatomically based model of the human body for switched gradient magnetic fields of a Maxwell pair of single-turn coils, each of diameter 0.654 m that are placed at an axial distance of 0.5664 m from each other. This is one of the typical geometries used for z-gradient coils for the present-day 1.5T MR imagers. |
||
| Quality Assurance Measures |
Our objectives, as stated in the proposal are to: 1. Extend the impedance method to higher frequencies by including inductances, and then to validate this formulation by comparing impedance-method solutions with known analytical solutions for configurations such as solid and layered spheres. 2. Evaluate the relative merits of the generalized impedance method compared to the modified FDTD method for calculating SARs produced by RF magnetic fields in MRI systems. 3. Calculate the SARs using both methods and compare results to establish the validity of the methods. Then calculate the SARs produced by MRI systems in our new high-resolution anatomically based models of the human body. |
||
| Results and Discussion |
Calculation of SARs for Body Coil: The SAR calculations have been done for the presently used
radio frequency of 64 MHZ for 1.5T MR imagers and for higher frequencies up to 170 MHZ that
may be used in the future. This has allowed us to obtain the peak SARs for any 1 kg of tissue for
which upper limits have been suggested in the NRPB safety guidelines.
Calculations of SARs for the Presently Used Head Coil: Using the finite-difference time-domain method, we have calculated the magnetic fields in the center of the head coil when it is empty or filled with 2/3-muscle-equivalent cylinder of diameter 21 cm. A cell size of 0.655 cm is used for the calculations. A 0.655-cm (nominal 1/4") resolution model of the human head, neck, and upper torso is used to calculate the SAR distributions not only at the presently used radio frequency of 64 MHZ for 1.5 T MR imagers, but also at higher frequencies up to 170 MHZ that may be used in the future. For all of these calculations we have assumed an outer shield of diameter 65.5 cm. Calculations of SARs for Proposed High-Frequency Head Coils: We have used physical dimensions typical of such coils and have calculated the SAR distributions of the 0.655-cm (nominal 1/4") resolution anatomically based model of the head, neck and upper torso for frequencies up to 400 Mz that are planned for these coils. Because of the resonant nature of the coil, a half cosine-wave-type current variation is assumed along the length of the rungs. Also, a 22.5 progressive shift between adjacent rungs is assumed to obtain circularly polarized RF magnetic fields. Currents Induced by Switched Gradient Magnetic Fields. Using the impedance method we have calculated the induced current density distributions in the 0.655-cm-resolution model of the human body for which anisotropic conductivities have been taken for heart and skeletal muscle. In order to increase the resolution of the model to dimensions on the order of 2-3 mm to properly model the fat or bone to wet tissue interfaces, it is important to truncate the model to regions of high current densities which correspond to the regions of the body where high values of dB/dt are created. We have verified that because the induced currents are primarily circulatory in nature, it si possible to truncate the models to focus on the high-current density regions. This approach results in considerable savings in computer memory and computation times. The highest induced current densities have thus been calculated for the various pulse sequences that at are typically used in present-day imagers. |
||
| Recent Publications |
O.P. Gandhi, S.B. Chen, X. Yuan and J.Y. Chen. Dosimetry of Time-Varying Magnetic Fields in
MR Imaging. Abstracts of Papers presented at the Sixteenth Annual Meeting of the
Bioelectromagnetics Society, Copenhagen, Denmark, June 12-17, 1994. p. 29.
D.N. Buechler, C.H. Durney, and O.P. Gandhi, 3D-Equivalent-Circuit (Impedance) Method for Solving Maxwell’s Equations. Abstracts of Papers presented at the Sixteenth Annual Meeting of the Bioelectromagnetics Society, Copenhagen, Denmark, June 12-17, 1994. p. 18. |
||