Yukio Kaneko1, Kosuke Ito2, Masahiro Takizawa2, Yoshihisa Soutome1,2, Hideta Habara1,2, Yusuke Seki1, Tetsuhiko Takahashi2, Yoshitaka Bito2, and Hisaaki Ochi1
1Research and Development Group, Hitachi Ltd., Tokyo, Japan, 2Healthcare Company, Hitachi, Ltd., Chiba, Japan
Synopsis
The B1+
inhomogeneity in a human body increases as the strength of a static magnetic
field increases. Previous studies showed the effect of the number of RF
transmit channels in RF shimming.
However, the effect for a
partial region of the lumbar spine in a sagittal plane has not yet been
investigated. In this study, the effect of the number of RF transmit channels
for regional RF shimming in the lumbar spine region was investigated. The
results show that 4-channel RF shimming can contribute to improving B1+
homogeneity and reducing the transmit RF power more than 2-channel RF shimming.Purpose
As the strength of a static magnetic field increases,
the B
1+ inhomogeneity in a human body increases. Various kinds
of multi-channel RF transmission techniques for reducing B
1+
inhomogeneity have recently been developed. One of these methods, RF shimming
1,2, is currently used in commercial MRI systems. The effect of the number
of RF transmit channels (ch) has been investigated numerically and
experimentally
3-6. However, the effect for lumbar spine (L-spine) imaging,
especially for the partial region in a sagittal plane, has not yet been
investigated. In this study, the effect of the number of RF transmit channels
between 2ch and 4ch regional RF shimming for L-spine imaging was investigated,
using numerical simulation. It was shown that 4ch RF shimming can contribute to
improving B
1+ homogeneity and reducing the transmit RF
power more than 2ch RF shimming.
Methods
The
effect of the number of RF transmit channels was confirmed using numerical simulation
of an electromagnetic field. The spatial distributions of B1+ in the human body were calculated using an electromagnetic
simulation tool (CST STUDIO SUITETM). Figure 1 shows the simulation model. A four-channel coil was
used for RF transmission7. The inner bore size (x-y plane) was 74 x 65 cm. Hugo
(height: 180 cm, weight: 90.3 kg) was used as a human model. Figure 2 shows the schematic of the setting of the Region
of Interest (ROI) used for RF shimming. The whole ROI covered the entire abdominal
region, and the partial ROI covered the back side where the L-spine exists. B1+ homogeneity was optimized, using B1+ map data in each region. The value of B1+
inhomogeneity (USD) was defined as below:
$$U_{SD}=\frac{\sigma}{\overline{B_{1}^{+}}},$$
where $$$\sigma$$$ is the standard deviation of B1+,
and $$$\overline{B_{1}^{+}}$$$ is the average of B1+
in a sagittal slice at x = 0 mm. The
normalized value of the total RF transmit power (PSUM) was defined
as below:
$$P_{SUM}=\frac{\sum_i^N|x_{i}|^{2}}{P_{SUM_{QD}}},$$
where $$$ x_{i}$$$ represents
the RF parameter, $$$i$$$ represents the
channel number, $$$N$$$ represents the
number of the channels (2 or 4), and $$$P_{SUM_{QD}}$$$
represents the total power in the case of quadrature (QD) drive.
The RF transmission mode was QD / 2ch RF shimming / 4ch RF shimming. 2ch RF shimming was conducted by combining B1+ maps of two channels (ch1 and ch3 / ch2 and ch4). 4ch
RF shimming was conducted by using a B1+ map of each channel. The USD or PSUM
was minimized in the RF shimming algorithm. The maximum local SAR was defined
as the maximum value of 10g SAR in the body, and the values were calculated for
the cases of RF shimming.
Results
Figure 3 shows the U
SD
and B
1+ maps in the case of RF shimming (minimization of
U
SD). The $$$\overline{B_{1}^{+}}$$$ in each ROI was normalized to 1 μT. The U
SD in
the case of 4ch RF shimming is smaller than that in the case of 2ch RF
shimming. Figure 4 shows the P
SUM and B
1+ maps
in the case of RF shimming (minimization of P
SUM, maintaining the U
SD
and $$$\overline{B_{1}^{+}}$$$ in the case of QD). The P
SUM in the
case of 2ch RF shimming is almost the same as that in the case of QD. On the
other hand, the P
SUM in the case of 4ch RF shimming is less than
that in the case of QD or 2ch RF shimming. Figure 5 shows the maximum local SAR
in the case of RF shimming for the partial ROI. The maximum local SAR in the
case of 4ch RF shimming is almost the same or less than those in the cases of
QD and 2ch RF shimming.
Discussion
In the case of 4ch RF shimming in Fig. 4, the higher
the reduction ratio of the P
SUM is, the narrower the setting of the ROI
is. The spatially asymmetric B
1+ field can be created by
using a four-channel transmit coil in Fig 4 (d). This is considered to
contribute to the reduction of transmit RF power. In Fig. 5, it is indicated
that the regional RF shimming method shown in this study can reduce U
SD
and P
SUM, not exceeding the value of the maximum local SAR in QD.
Conclusion
The effect of the number
of RF transmit channels for RF shimming for the partial region in L-spine
imaging has been investigated using numerical simulation. It is shown that 4ch RF
shimming can contribute to reducing the B
1+ inhomogeneity
and the transmit RF power more than 2ch RF shimming.
Acknowledgements
No acknowledgement found.References
[1] Nistler J, Diehl D, Renz W, et al. ISMRM
2007; 15: 1063.
[2] Hajnal JV, Malik SJ, Larkman DJ, et
al. ISMRM
2008; 16: 496.
[3] Childs AS, Malik SJ, O’Regan DP, et
al. MAGMA 2013; 26(4): 401-410.
[4] Harvey PR, Possanzini C, Simons J. ISMRM 2010; 18: 1486.
[5] Kaneko Y,
Soutome Y, Habara H, et al. ISMRM 2013; 21: 2756.
[6] Ito K, Kaneko
Y, Soutome Y, et al. ISMRM
2014; 22: 942.
[7] Soutome Y,
Otake Y, Dohata M, et al. ISMRM 2013; 21: 2750.