de Oliveira

Document Sample
de Oliveira Powered By Docstoc
					                                                                               39. DGMP Tagung 2008 in Oldenburg

A New Inductively Coupled Coil to Improve Local Signal-to-
    Noise Ratio during MR-guided Prostate Biopsies
de Oliveira, André; Umathum, Reiner; Krafft, Axel; Semmler, Wolfhard; Bock, Michael
Deutsches Krebsforschungszentrum (dkfz), Abt. Medizinische Physik in der Radiologie,
Im Neuenheimer Feld 280, 69120 Heidelberg
It has recently been shown that magnetic resonance spectroscopy (MRS) can be used to improve specificity in
the detection of prostate cancer [1]. During MR-guided prostate biopsies the MRS metabolic information might
help to decide from which areas biopsy samples need to be harvested. Since the quality of MRS crucially de-
pends on the local signal-to-noise ratio (SNR), we propose a new inductively coupled coil that can be integrated
into an MR-compatible biopsy system. The coil provides a higher local SNR in the range of the biopsy needle.
Material and Methods
Prostate Biopsy Device
For the MR-guided prostate biopsies a commercially available prostate biopsy device [2] (Invivo, Germany) is
used that consists of a plastic cylinder filled with a contrast agent solution which is co-linear with the biopsy
needle. This passive marker is connected to a holder system that allows movements in 3 degrees of freedom to
target the prostate (Fig. 1).
Inductively Coupled Coil Design
To amplify the MR signal near the tip of the biopsy needle, an inductively coupled RF coil was mounted in a
toroidal Plexiglas coil former which was tightly attached to the passive marker. The solenoid coil had an inner
diameter of 23.0 mm, with 3 turns of copper wire (diameter = 0.2 mm), an inductance of 567 mH and a series
resistance of 1.7 Ω. A serial capacitor (2-12pF) was used to tune the coil’s resonance frequency to the Larmor
frequency (63.7 MHz at 1.5 T) of the MR system. A pair of anti-parallel diodes (BA 792, Philips
Semiconductors, Netherlands) was used to passively detune the coil during RF transmission of the MR scanner.
MR Signal Amplification
A phantom (contrast agent solution Gd-DTPA/water: 1:300) was used to determine the coils sensitivity profile
along its symmetry axis. The coil was orientated perpendicular to the B0 field and a FLASH image (TR = 9.5 ms,
TE = 5.0 ms) was acquired. The MR signal along the coil symmetry axis was measured and normalized to the
average signal more than 5 coil radii away from the coil (i.e., outside the sensitivity region of the coil).
B1 Field Homogeneity
Due to the inductive coupling principle, B1 inhomogeneity can result in the vicinity of the coil, even though a
crossed diode circuitry was inserted to minimize coupling during RF transmission. The same phantom was used
to quantify the coil’s effect on the B1 homogeneity of the MR system. The inductively coupled coil was
orientated perpendicular to B1 field and FLASH images (TR = 9.5 ms, TE = 5.0 ms) were acquired with the flip
angle varying in the range 1°-45°. The signal was plotted as a function of the flip-angle and a non-linear fit to
the FLASH equation was performed. The Ernst
angle  was obtained from the maximum of the
signal, and the B1 amplification at each position
(x,y) was calculated as ,norm / E(x,y). Here,
,norm denotes the Ernst angle far away from the
MR Spectroscopy
A MRS phantom containing a solution of water,
choline (40 mM), creatine (16 mM) and citrate
(3mM) was built to determine SNR
improvements in MRS provided by the
inductively coupled coil. A single voxel spin
echo     spectroscopy     acquisition   (TR/TE
=1500/135 ms, repetitions = 64) was repeated 10       Fig. 1: MR-compatible prostate biopsy passive marker holder
                                                      showing the three degrees of freedom (a,b). The inductively
times with and without the coil in place. The
                                                      coupled coil (c) is attached to the prostate biopsy passive
voxel was placed 14 mm away from the biopsy           marker (d).
coil center (expected distance between the coil

Fußzeile TimesNewRoman 10
                                                                                                            39. DGMP Tagung 2008 in Oldenburg

center and the biopsy target). The SNR calculation
was done as in [3] by taking the ratio between the
choline/creatine/citrate peaks and the noise
standard variation in the flat baseline region (>6 or
<0 ppm). The SNR improvements were quantified
as the ratio between the SNR with and without the
The signal gain and B1 homogeneity are shown in            Fig. 2: B1 homogeneity map showing coil’s position.
Fig. 3. At 14 mm, the signal gain and the B1
amplification were 90% and 17%, respectively.
Figure 3 shows the B1 homogeneity map. In the
regions near the coil center, a deviation of up to                                                                               B1 amplification
25% was observed whereas the B1 field was                                                                                        Coil amplification
homogenous in regions further than 30 mm from
the coil. The SNR results for choline, creatine and
citrate with/without the coil were 11.7±0.8 /
6.6±1.8, 6.0±0.8 / 3.2±1.1, and 2.7±0.9 / 1.8±0.8,
respectively. The SNR improvements for choline,
creatine and citrate were 1.8±0.5, 1.8±0.7 and
1.5±0.8. Figure 4 shows a typical spectrum
measured in this experiment.
Discussion and Conclusion
The proposed coil concept provides a nearly                                          0      10             20              30          40              50   60
twofold sensitivity increase in the target area of                                                    Distance from the coil [mm]
the biopsy. It is especially designed for com-               Fig. 3: Graph showing coil characterization: signal gain and
bination with a commercial prostate biopsy                   B1 amplification as a function of the distance from the coil
system, where the expected distance between coil             center.
                                                                                                 Citrate        Creatine    Choline
and center of the biopsy region is about 14 mm,                                      6000
and only a minimal B1 amplification was seen. In
                                                                                                                                            Without coil
this system, the inclination between the coil and B0                                 5000                                                   With coil
direction is not expected to be smaller than 30°,
which would be related with a signal loss of about                                   4000
                                                              Intensity (a.u.)

50% in the worst case.
During MR-guided prostate biopsies, targets are            3000

often defined as low intensity regions visible in
T2w images. Although previous information from             2000

diagnostic imaging is available, we believe that the
additional information provided by a current MRS           1000
measurement could help to identify suspect
regions. The fact that the inductively coupled coil           0
does not need any additional cables to be connect              2,0      2,5      3,0         3,5 8,0               10,0

to the MR-scanner together with its low                                         Frequency shift (ppm)
manufacturing costs may present the ideal solution       Fig. 4 :Graph showing the measured MR spectrum of the MRS
for improving local SNR during MR-guided                 phantom with and without the coil in place. The range between
prostate biopsies. The integration of the proposed       8,0 and 10,0 ppm shows part of the flat baseline region.
inductively coupled coil and the prostate biopsy
device may represent a step forward in prostate cancer diagnosis performance. The observed SNR gains together
with recently proposed passive tracking systems [4] may help increasing MR-guided prostate biopsies sensitivity
and specificity without increasing procedure duration.
[1] Swindle P, McCredie S, et al. Pathologic characterization of the human prostate tissue with proton MR spectroscopy.
Radiology 228:144-151 (2003).
[2] Beyersdorff D, Winkel A, et al. MR imaging-guided prostate biopsy with a closed MR unit at 1.5 T: initial results.
Radiology 234:576-581 (2005).
[3] Bartekka K, Sunitha B, et al. Enhancing nonmass lesions in breast: Evaluation with proton (H) MR spectroscopy.
Radiology 245:80-87 (2007).
[4] de Oliveira A, Rauschenberg J, et al. Automatic passive tracking of an endorectal biopsy device using phase-only cross
correlation. Magnetic Resonance in Medicine 59:1043-1050 (2008).

Fußzeile TimesNewRoman 10

Shared By: