Fig 1.
a)–Full-body Visible Human Project Female v5.0 CAD based computational phantom. The phantom is composed of 249 individual structures. Some individual muscles are removed for clarity. b-e) Examples of coregistration maps for surface meshes in different transverse planes with tissue labeling. In case b), non-anatomical separation between scalp and skull was corrected. In case d), the tissue labeling list is not shown. The complete full-body coregistration maps with the vertical resolution of 1 mm and with tissue labeling in every cross-sectional plane are available online in *.mp4 format [24] for independent inspection and verification purposes.
Fig 2.
Local SAR distribution in the coronal plane for a high pass full-body RF coil operating at 64 MHz loaded with the VHP-Female v5.0 computational phantom given amplitude of 1 μT at the coil center.
a)–Positioning of model within the birdcage at shoulder/heart landmark. b)–Ansys HFSS (Electronics Desktop) solution with one adaptive pass. c)–Solution with eight adaptive passes.
Fig 3.
Simulations to establish calibration for experiments conducted in [43–45].
a)–The metallic nail implant placed within a homogeneous loop-like Agar phantom [43–45] at a depth of 2 cm. b)–The current density produced with a whole-phantom SAR of 4.0 W/kg. c)–The simulated temperature given a total volumetric power loss of 120 W exactly corresponding to experiment [43–45]. Simulation results produced a temperature rise of 10.02°C, slightly less than the 12.6°C experimentally observed in [43–45].
Fig 4.
The VHP-Female computational phantom positioned with a 1.5 T MRI birdcage coil.
Some body parts are removed for clarity. a)–The femur position is illustrated to show its orientation within the model–the metallic nail implant is aligned to reside within the trabecular bone structure. b)–Each individual object within Ansys Mechanical model is assigned specific thermal properties. c)–The temperature rise is shown after 900 seconds of continuous coil operation. These values correspond well with published experimental data [43–45]–see Table 2.
Fig 5.
Comparison of the simulated numerical values (red stars) obtained using the VHP Female computational phantom with published experimental data (two black curves) [43–45] for the maximum temperature rise near the implant.
The depth of the implant within the VHP model is approximately 4 cm and the maximum temperature value after 900 seconds of coil operation time is represented by the top red star. The VHP model predicts a slightly higher (by 2.75°) maximum temperature rise than what was measured in [43–45]–see Table 2. This is likely due to the different (non-homogeneous) material properties employed in the present study and the slightly angled orientation of the nail.
Fig 6.
The VHP-Female computational phantom positioned with a 3T MRI birdcage coil at the abdominal landmark.
a)–The model geometry is shown, providing a view of the phantom located in the coil. b)–The simulated map generated in the axial plane about the model mid-section. c)–The experimental
map reported in [46,47]. d)–The simulated
map generated in the coronal plane about the VHP Female model mid-section. e)–The measured
map generated in a coronal plane at the approximate middle of the model. These values are also reviewed in Table 2.
Fig 7.
a,b)–The forearm of the VHP-Female computational human model adjacent to a single loop coil antenna operating at 165 MHz. The model geometry is shown, including internal forearm tissues, each defined with specific electromagnetic and thermal properties. c,d)–Computed (VHP-Female) and measured [49] local SAR within a plane directly above the center of the coil, respectively. Note the perfect agreement in the peak SAR. Also note the higher values generated due to the presence of the veins and nerves. e,f)–Computed (VHP-Female) and measured [49] temperature rise within a plane directly above the center of the coil, respectively, after 139 seconds. The maximum temperature difference is 1.4°C.
Table 1.
Comparison of SAR values predicted using the VHP-Female surface CAD model with the values predicted by a 5 mm voxel model derived from the identical image dataset [71].
Nearly identical high-pass birdcage coil dimensions, coil landmark, and excitation type were used. All results are normalized to 1 μT field at the coil center.
Table 2.
Comparison of simulated numerical results using the VHP Female computational model with experimental results for examples 2, 3, and 4.