Research Article1198791
Relaxation Measurement of Ex-Vivo Breast Cancer Tissues atUltralow Magnetic Fields
Seong-Joo Lee1 Jeong Hyun Shim1 Kiwoong Kim12 Seong-min Hwang1
Kwon Kyu Yu1 Sanghyun Lim12 Jae Ho Han3 Hyunee Yim3 Jang-Hee Kim3
Yong Sik Jung4 and Ku Sang Kim4
1Center for Biosignals Korea Research Institute of Standards and Science (KRISS) 267 Gajeong-ro Yuseong-guDaejeon 305-340 Republic of Korea2Department of Medical Physics University of Science and Technology (UST) 217 Gajeong-ro Yuseong-guDaejeon 305-333 Republic of Korea3Department of Pathology Ajou University School of Medicine 164 World Cup-ro Yeongtong-gu Suwon 443-380 Republic of Korea4Department of Surgery Ajou University School of Medicine 164 World Cup-ro Yeongtong-gu Suwon 443-380 Republic of Korea
Correspondence should be addressed to Kiwoong Kim kwkimkrissrekr
Received 22 August 2014 Revised 15 October 2014 Accepted 21 October 2014
Academic Editor Enzo Terreno
Copyright copy 2015 Seong-Joo Lee et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
We investigated 1198791relaxations of ex-vivo cancer tissues at low magnetic fields in order to check the possibility of achieving a 119879
1
contrast higher than those obtained at high fieldsThe1198791relaxations of fifteen pairs (normal and cancerous) of breast tissue samples
were measured at three magnetic fields 37 62 and 122 120583T using our superconducting quantum interference device-based ultralowfield nuclear magnetic resonance setup optimally developed for ex-vivo tissue studies A signal reconstruction based on Bayesianstatistics for noise reduction was exploited to overcome the low signal-to-noise ratio The ductal and lobular-type tissues did notexhibit meaningful 119879
1contrast values between normal and cancerous tissues at the three different fields On the other hand an
enhanced 1198791contrast was obtained for the mucinous cancer tissue
1 Introduction
Breast cancer is one of the leading causes of cancer mortalityin women Undoubtedly early diagnosis of breast canceris beneficial to patients in terms of lowering mortality andenabling breast-conserving surgery The mammogram iswidely used for screening however it has the limitation oflow sensitivity to dense breast tissue [1] and the methodof breast compression necessary for this procedure mayresult in patients experiencing discomfort and pain Inparticular a significant population in some Asian countrieshas insufficient breast volume to respond to mammographicexamination properly and additionally these patients feeluneasywith this examination as a result of cultural reluctanceIn addition to mammograms magnetic resonance imaging(MRI) is considered a supplementary scanning technique butits role in screening is debated [2 3] The American Cancer
Society recommends MRIs only for people who have a highlife-time risk (gt20) due to family history [4] In order toobtain an improvedmagnetic resonance (MR) image of breasttumors patients may have to take contrast agents whichin principle are preferably avoided because of potential sideeffects Therefore the development of a new modality thatprovides confirmative imaging for early diagnosis of breastcancer still needs to be pursued in biomedical research
In general the discrimination between tumors andhealthy tissue usingMRI primarily relies on differences in thewater density of these tissues In the early stages of cancerhowever the water density in the cancerous tissue is expectedto be similar to that of normal tissue To achieve a 119879
1-
contrast MR image of the cancerous tissues magnetic fieldstrength should be adjusted to elaborate the difference in thespin-lattice relaxation times (119879
1) Field-cycling relaxometry
studies [5ndash8] have been performed on various biological
Hindawi Publishing CorporationBioMed Research InternationalVolume 2015 Article ID 385428 9 pageshttpdxdoiorg1011552015385428
2 BioMed Research International
mediums the human brain have shown that the largest1198791contrast is accessible at low magnetic fields that is at
less than 300 120583T [9] An agarose gel solution has showndrastically enhanced 119879
1contrasts at magnetic fields of less
than 100 120583T [10 11]The1198791-weighted imaging of the tissue of a
rat tumor was obtained at approximately 100 120583T [12] Recent1198791measurements of prostate tissue at 132120583T have revealed
that the 1198791contrast has a linear dependence on the ratio
of the cancer volume in the tissue samples [13] Inspired bythese previous studies we focus on ex-vivo relaxometry ofbreast cancer tissue samples at low magnetic field conditionshere Since the mammary gland is functionally and quali-tatively similar to prostate tissue an enhanced 119879
1contrast
of breast cancer tissue can be expected at such low fieldsOur experimental procedure is conducted on a homebuiltultralow field (ULF)-nuclear magnetic resonance (NMR)setup [14ndash19] using superconducting quantum interferencedevice (SQUID) to sense weakNMR signals owing to the lowstatic magnetic field
SQUID-based microtesla NMR has paved the way fornew biomedical measurements under low magnetic fieldconditions for example chemical analysis from a 119869-couplingspectrum [20] simultaneous proton density imaging withmagnetoencephalography [21 22] direct neutral currentimaging [23ndash25] brainwave magnetic resonance [18] andheartmagnetic resonance [26] All of thesemeasurements aredifficult to achieve with conventional high-field NMRMRIsThe present studyrsquos aim to achieve confirmative MR imagingof cancer tissues in-vivo and contrast agent free will certainlybe one of the most fruitful applications of low field MRI
In this study we measured the 1198791contrasts of breast
cancer tissues at three different low fields 37 62 and122120583T The 119879
1trend variation may give us a relaxometric
fingerprint to discriminate between breast cancer and healthytissue either by estimating an optimal magnetic field or bycomparing the slope of the 119879
1variation along the different
external magnetic fields Since the tissue volumes were small(approximately 1 cm3) relatively low signal-to-noise ratios(SNR) were obtained Hence we used a signal reconstructionmethod based on Bayesian statistics for noise removal andreliably separated signal amplitudes from noise Accordingto the model for relaxation in a multiphase system [27]provided that the exchange rate is faster than the relaxationrate (1119879
1) the relaxation rate has a linear dependence on the
volume ratio and the coefficient of the linearity provides the1198791contrast between themedia From the same analysis of our
data the 1198791contrasts at three magnetic fields failed to exhibit
ameaningful relevance to the volume ratio of breast cancer intissue One result worth noting is that mucinous carcinomatissue can be distinguished clearly from other tissue becauseof its high water density and long 119879
1relaxation time Possible
implications of our results and further studywill be discussed
2 Methods
21 Sample Treatment In this study fifteen pairs of breasttissue specimens were investigated Each pair consists of
nominal tumor and normal tissues All the tissue sam-ples were obtained from total mastectomy specimens inAjou University Hospital (Suwon Korea) The tumor tissuesamples were found to include various cancer types teninvasive ductal carcinomas one ductal carcinoma in situone mixed ductal and lobular carcinoma and one mucinouscarcinoma Apart from the mucinous case discriminationbetween cancer types was not the concern of this study Thisresearch was approved by the Institutional Review Board ofAjou University Hospital
After gross examination of each mastectomy specimenrepresentative tumor and normal tissue samples with vol-umes of approximately 1 cm3 were prepared and then imme-diately inserted into a liquid nitrogen (LN2) tank where theywere stored until the NMR experiments were conductedfollowing Huang et al [8] Prior to the 119879
1measurements
the tissues were thawed in the ambient atmosphere andtemperature To avoid dehydration after thawing they werewrapped in cling film
The specimens were formalin-fixed after the NMR mea-surements were complete The paired tumor and normaltissue samples were routinely processed and stained withhematoxylin-eosin for histologic examination From eachtissue a single slide was prepared via sectioning and exam-ined with a microscope The proportion ratios of cancer cellswere measured by mapping the slides on a square grid inmillimeter
22 Experimental Setup Figure 1 shows the experimentalsetup for the microtesla NMR experiment All equipmentwas mounted inside a magnetically shielded room (MSR)which was specially designed for ULF-NMRMRI [18] A dc-SQUID (SupraconAG)was used as aNMR signal detector forthe ULF-NMR system The dc-SQUID sensor was shieldedwith two different cans The inner shield is a superconduct-ing Nb cast can with 999 purity The outer shield is asuperconducting lead can with 995 purity The pickup coilis a second-order gradiometer which is made of a 125120583mNb wire with a 50mm baseline and a diameter of 65mmThe dc-SQUID with pickup coil was additionally wrappedwith an aluminum-coated Mylar film to prevent the SQUIDsystem from being influenced by ambient RF noise Theenvironmental noise of our system inside the MSRmeasuredby the dc-SQUID sensor with second-order gradiometricpickup coil is about 23 fTHz12 at 100Hz
Since magnetization in the microtesla region is insuffi-cient to produce the NMR signal in the ULF-NMR a strongmagnetic field (119861
119901) for prepolarizing the nuclear magnetiza-
tion must be applied prior to NMR signal acquisitionThe 119861119901
strength is in tens of mT The free precession of the nuclearmagnetization is then recorded under a measurement field(119861119898) in the 120583T region A double Helmholtz coil [28] was
used to produce a high degree of homogeneity in the 119861119898
The119861119898strength at the sample space was about 493120583T which
corresponds to a proton NMR frequency of about 210HzProducing a high 119861
119901leads to the heating of the 119861
119901coil
due to the injection of a high current For the measurementof the ex-vivo tissues the temperature increase in the tis-sue caused by heat transfer from the 119861
119901coil may induce
BioMed Research International 3
SQUID dewar
Reservoir
N2 gas
LN2
1st Bp2nd Bp
Bm
x
y
z
dewar
Figure 1 Schematic diagram of the experimental apparatus for themicrotesla NMR experimentThe axes of the 1st and 2nd 119861
119901coils are
parallel However these are perpendicular to the axis of the 119861119898coil
The 1st 119861119901coil is a liquid nitrogen (LN2)-cooled pancake-type coil
The LN2 flowed continuously from the reservoir to the 1st 119861119901dewar
through two tubes at the bottom and then the nitrogen gas flowedback to the reservoir through the upper tube
undesired effects such as tissue degeneration and variationin 1198791relaxation time It is therefore important to maintain
a constant tissue temperature during the measurements The119861119901coil system corresponding to the 1st 119861
119901dewar in Figure 1
was designed for this purposeFor effective generation of the 119861
119901at the sample region
which is beneath the base of the SQUID dewar we adopted apancake-type geometry for the 119861
119901coil A 700-turn copper-
wire-wound pancake coil which has an outer diameter of132mm and a length of 16mm generates a magnetic field of60mT at the region of the sample with a current injection of20A
The 119861119901coil was mounted in a dewar which is the green
cylindrical component labeled ldquo1st 119861119901dewarrdquo in Figure 1 for
effective cooling Liquid nitrogen (ldquoLN2rdquo in Figure 1) flowscontinuously from the reservoir to cool the 119861
119901coil Nitrogen
gas produced by the resultant boiled liquid then flows backout to the reservoir The multiple layers of glass fiber inthe covering plate of the dewar prevent the refrigerationof the tissue A 3D-printed 4-mm thick thermal shieldplate having a meandering air path was additionally placedbetween the sample and the covering plate of the 119861
119901dewar
in order to keep the tissue temperature constant against therelatively cold surface of the dewarTheN
2gas was channeled
continuously through the thermal shield plate resulting in atemperature variation of plusmn03∘C for thirty minutes on the topsurface of the plate
A Helmholtz coil was used for the 2nd 119861119901 The field
strengths were varied between 37 62 and 122 120583T whichwere measured at the sample space The 119861
119898and the 2nd 119861
119901
coils were controlled by solid state relays (SSR) Howevera large amount of power is necessary to generate 1st 119861
119901of
approximately 60mT which should be switched off after a
Signal
1st Bp
2nd Bp
Bm
75ms 10ms
tB119901 tdelay tmeasurement trepetition
Figure 2 Illustration of pulse sequence for1198791experimentsThe data
set of 1198791is obtained by changing the duration of 119905delay
few milliseconds Therefore the 1st 119861119901coil was controlled
by a specially constructed current-driving circuit [17] whichis basically composed of a capacitor bank and an insulatedgate bipolar transistor [29]The SSRs and the current-drivingcircuit were remotely switched by a programmable pulse-generating board connected by optical fibers to protect theULF-NMR system from external electronic noise
Figure 2 shows the pulse sequence for the 1198791experiment
Initially the 1st 119861119901is applied to form net magnetization
in one direction After the 1st 119861119901is turned off the spins
begin to relax while the 2nd 119861119901is turned on After the
time 119905delay has elapsed the spins produce a free precessiondecay (FPD) signal which is measured by the SQUID sensorSince the degree of relaxation differs in accordance with theduration of 119905delay a 1198791 curve is obtained by varying 119905delay Thefollowing experimental parameters were used in this study(see Figure 2) 119905
119861119901= 1 s 119905measurement = 1 s and 119905repetition = 7 s
The averaged FPD signals were measured at each time 119905delayThe number of averages was mainly 20sim40 depending on thesample We measured mainly 4 points with different 119905delayvalues for each sample in order to generate the 119879
1curve
23 Analysis MethodmdashBayesian Analysis ANMR signal canbe expressed as the summation of 119873 exponentially decayingsinusoids and Gaussian noise [30]
119889119896=
119873
sum
119895=1
[(119860119895119890119894120601119895) 119890minus119896Δ119905120591119895
1198902120587119894119896Δ119905119891119895
] + 120576119896
119896 = 0 119872 minus 1
(1)
where 119860119895is the signal amplitude 120601
119895is the phase 120591
119895is the
decay time 119891119895is the frequency of the 119895th sinusoid119872 is the
number of data points and 120576119896is Gaussian random noise with
zero mean and standard deviation 120590 Since this informationabout the NMR signal is intensively located in the beginningpart of the FPD signal it is therefore important to record theinitial part of that signal
If the SNR is low even the initial signal is already hiddenby noise as was the case for the tissue samples investigatedin this study Therefore we adopted the Bayesian analysismethod [30] which is suitable for the analysis of small signalsThe primary assumption is that the noise obtained during thedata acquisition has a Gaussian distribution After adopting
4 BioMed Research International
the Bayesian rule the probability becomes a function of thedecay time and the resonance frequency Consider
119875 (d | 120591 f 120590119873)
=
(21205871205902)119873minus119872
1205821sdot sdot sdot 120582119873
times exp[[
[
minussum119872minus1
119896=0
10038161003816100381610038161198891198961003816100381610038161003816
2
+ sum119873
119895=1120582119895
100381610038161003816100381610038161003816(ddaggerUE)
lowast
119895120582119895
100381610038161003816100381610038161003816
2
21205902
]]
]
(2)
where U = exp(minus119896Δ119905120591119895) exp(2120587119894119896Δ119905119891
119895) and E and 120582 are
the matrices of the eigenvectors and eigenvalues of UdaggerUrespectively The equation of probability in [30] needs to beadjusted as illustrated in (2) Therefore the solution is to findthe maximum probability by varying two variables the decaytime and the resonance frequency When the probability isat a maximum the signal intensity and the phase are alsodetermined as
119860119895119890119894120601119895=
[E (ddaggerUE)lowast
]119895
120582119895
(3)
Since the resonance frequency can be determined by theFFT of the relatively larger signal or the measurement ofthe 119861119898strength the variable of the resonance frequency is
additionally reducedTherefore more precise information onthe signal intensity could be obtained from the FPD signalDuring the analysis of the acquired data nondecaying termssuch as the higher harmonics of the AC line noise andorthe additional noise were also considered by adding theseterms to the equations whereas only one decaying termcorresponding to the FPD signal was used
3 Results
To estimate 1198791relaxation times the variation in the FPD
signal amplitudes was measured as a function of the timeduration 119905delay Each data set was fitted using the followingequation
119891 (119905) = 120572 exp(minus119905delay
1198791
) (4)
where 120572 is the amplitude and 1198791is the longitudinal relaxation
timeThe additional constant to represent the base-offset wasnot included because the Bayesian analysis described abovediscriminates NMR signals from noisy backgrounds well(see Figure 3 andor Supplementary Information in Supple-mentaryMaterial available online at httpdxdoiorg1011552014385428)
Table 1 summarizes the pathologic characteristics frac-tion of cancerous cells in each tumor and normal tissue (therest of the tissue is filled with normal cells) and themeasured1198791values at three magnetic fields of the thirteen pairs of
breast tissue specimens which exhibited FPD signals The
Original
00 01 02 03
Time (s)
1
0
minus1
FPD
sign
al (a
u)
(a)
With noise
00 01 02 03
Time (s)
3
0
minus3
FPD
sign
al (a
u)
(b)
Reconstructed
00 01 02 03
Time (s)
1
0
minus1FP
D si
gnal
(au
)
(c)
Figure 3 Simulation of Bayesian analysis (a) the simulated FPDsignal (b) the mixed signal with Gaussian random noise and (c)the reconstructed signal from the noisy signal plotted in (b) Thesimulated FPD signal was built on assumptions of 119860
119895= 1 120601
119895= 0
120591119895= 008 and 119891
119895= 211 in (1) The nondecaying signals were also
added as electronic noise which had the same amplitude of 002and frequencies of 180 and 240Hz The simulated FPD signal wasmixed with Gaussian random noise which had a standard deviationof 08 After Bayesian analysis the reconstructed signal had thedecaying components 119860
119895= 1031 120601
119895= 0011 120591
119895= 0066 and
119891119895= 211 The nondecaying components of the reconstructed signal
were composed of the amplitudes and phases of 119860180
= 0026119860240= 0027 and 120601
180= minus0857 120601
240= 0314 respectively
examples of 1198791fitted curves are shown in the Supplementary
Information section In addition two 1198791values could not
be estimated using the fitting equation because the numberof data points obtained by the FPD signal were insufficientIn those cases the baseline of the FPD signal fluctuates andis discontinued Most tumor specimens were composed ofcancer cells surrounding desmoplastic stroma and somenormal epithelial cells For example the invasive ductalcarcinoma (Specimen number S13-10) plotted in Figure 4(a)shows irregularly infiltrating epithelia cell clusters and awell-organized normal breast lobule which are denoted bythe arrows on the right- and left-hand sides of the picturerespectively The mucinous carcinoma specimen howeverhad only cancer cells and mucin with no normal tissueThemucinous carcinoma sample (Specimen number S13-09)plotted in Figure 4(b) showed a few clusters of malignantepithelial cells in the background of the mucin pool whereasthe normal breast tissue was composed of fibrotic stroma anda few epithelial cells (see Figure 4(c)) Among the four cases
BioMed Research International 5
Table1Analyzedandmeasuredparameters
Specim
ennu
mber
Type
Cancer
1198791at122120583
T(m
s)1198791at62120583T(m
s)1198791at37120583T(m
s)Tu
mor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
S13-01
Ductal
900
604plusmn457
482plusmn264
841plusmn219
914plusmn422
1095plusmn429
69plusmn3
S13-02
aDuctal
250
1611plusmn532
813plusmn247
535plusmn108
941plusmn332
79plusmn352
564plusmn193
S13-03
Ductal
750
687plusmn99
74plusmn385
634plusmn29
623plusmn205
668plusmn137
977plusmn184
S13-04
aLo
bular
600
81plusmn201
546plusmn11
618plusmn5
608plusmn5
595plusmn89
66plusmn84
S13-05
aDuctal+
lobu
lar
605
1001plusmn
74mdash
121plusmn
533
866plusmn5
685plusmn104
105plusmn124
S13-06
Ductal
700
722plusmn228
673plusmn75
636plusmn138
676plusmn83
794plusmn52
634plusmn53
S13-07
Ductal
750
792plusmn58
1013plusmn112
736plusmn48
744plusmn123
682plusmn48
742plusmn64
S13-08
Ductal
900
1692plusmn147
833plusmn156
819plusmn283
829plusmn93
641plusmn184
612plusmn128
S13-09
Mucinou
s100
03477plusmn249
775plusmn269
3586plusmn258
847plusmn134
mdash46
3plusmn56
S13-10
Ductal
750
67plusmn3
635plusmn97
689plusmn27
844plusmn19
656plusmn24
659plusmn56
S13-11
aDuctal
900
861plusmn79
984plusmn298
84plusmn61
603plusmn44
3765plusmn13
876plusmn278
S13-12
Ductal
700
115plusmn117
1006plusmn85
912plusmn33
812plusmn27
498plusmn71
255plusmn78
S13-13
Ductal
900
655plusmn03
758plusmn212
522plusmn154
773plusmn81
64plusmn31
816plusmn145
a Thep
reop
erativec
hemotherapy
was
perfo
rmed
6 BioMed Research International
(a) (b)
(c)
Figure 4 Photomicrographs of (a) the invasive ductal carcinoma sample (Specimen number S13-10) (b) the mucinous carcinoma sample(Specimen number S13-09) and (c) normal breast tissue (Specimen number S13-09) The arrows on the right- and left-hand sides of thepicture in (a) represent irregularly infiltrating epithelia cell clusters and a well-organized normal breast lobule respectively Each specimenwas stained with hematoxylin-eosin and observed at times40 magnification
receiving preoperative chemotherapy one case (Specimennumber S13-02) responded to treatment and the majority ofthe tumor tissue was replaced with fibrotic stroma
Since the tumor tissues contain partial cancers the 1198791
values estimated from the fitting do not yield meaningfulinformation about the 119879
1contrast unless properly analyzed
For correct estimation of the 1198791contrast between cancer and
normal tissue the cancer volume ratio must be considered aspart of the analysis The model devised by Zimmerman andBrittin [27] describes the relaxation phenomenon in a systemof multiple phases According to this model the exchangerate between the twomedia with different119879
1relaxation times
strongly influences the 1198791relaxation process We presume
that the exchange rate of the water molecules betweennormal and cancer cells in a tumor tissue is reasonablyrapid compared with the relaxation rate 1119879
1 Then a single
exponential decay with the average relaxation rate 11198791119879=
1205881198621198791119862+ 1205881198731198791119873 will be obtained in which 119879
1119862and 119879
1119873
are the 1198791relaxation times of the cancer and normal cells
respectively The 120588119862and 120588
119873are volume ratios of the cancer
and normal cells in the tissueThe ratio between1198791119873
and1198791119879
that is 11987911198731198791119879 of each tumor tissue should show a linear
dependence on the 120588119862and the coefficient is the 119879
1contrast
120575 between the cancer and normal cells as
1198791119873
1198791119879
= 1 + 120588119862(1198791119873
1198791119862
minus 1) = 1 + 120588119862120575 (5)
Figure 5 shows the 1198791ratios between the tumor and the
healthy tissue (11987911198731198791119879) as a function of the relative cancer
volume 120588119862 The (red) closed circles (blue) open circles and
(green) closed squares represent the data obtained at 122 62and 37 120583T respectively Apparently the119879
1relaxation time has
a weak dependency on 120588119862 and the 119879
1contrast 120575 exhibits a
weak variance as the magnetic field increases The estimated120575 is 33 times 10minus4 at 37 120583T 63 times 10minus4 at 62120583T and minus94 times 10minus4 at122 120583T which do not appear to bemeaningful compared withstandard deviations of the data Possible reasons for this willbe discussed below
From the mucinous carcinoma tissue (Specimen numberS13-09) a clearly distinguishable NMR signal was measuredas shown in Figure 6 The NMR spectrum has a full-widthhalf-maximum (FWHM) of 17Hz which is about onequarter of the FWHMs of the other tissue which are typically7Hz As a result 120575 is higher than 4 As in the case of high-field MRI [31] mucinous carcinoma tissue can be identifiedunambiguously in low field MRI via 119879
1or 1198792weighted
imaging methods
4 Discussion
In the multiphases relaxation model [27] (5) is based onthe assumption that water molecules in cancer and normalcells can be exchanged rapidly compared with the relaxation
BioMed Research International 7
20 30 40 50 60 70 80 90
Ratio of cancer ()
2
1
0
T1
NT
1T
Figure 5 The 1198791ratio between the tumor (T) and the normal (N)
tissue as a function of the cancer ratio Since the normal tissue inspecimen S13-05 was 5 cancerous the surface ratio of the cancerof this specimen was denoted as 55 The (red) closed circles(blue) open circles and (green) closed squares represent the dataobtained at 122 62 and 37 120583T respectively The error bars indicatethe standard errors The lines are linear fits to data obtained at eachfield
190 200 210 220 230000
005
010
015
020
025
Nor
mal
ized
FFT
(au
)
Frequency (Hz)
FWHM 7Hz
FWHM 17Hz
Figure 6 The normalized spectra for the tumor tissues of S13-09(red) and S13-10 (green) obtained by the experimental conditions of119905delay = 10ms and 2nd 119861
119901= 62 120583T The spectra were normalized
with each spectrum area
rates of water molecules belonging to separate phases in thesystem If this assumption is not valid the 119879
1relaxation of
tissue should exhibitmultiple exponentialsThe1198791values dis-
played in Table 1 are mostly lt100msTherefore the exchangerate between water molecules in cancer and normal cellsshould be at least 10 sminus1 in order for (5) to be applied Suchan exchange rate might not be a realistic value We speculatethat cancer cells are still isolated from normal cells for longerperiods than the time scale of 100ms In this case multipleexponential fitting of the decay curve should be required tocorrectly analyze tissue samples having a cancer ratio in therange of 20sim80The relaxometry of a single (or finite) voxel
in the MR imaging of tumor tissue will therefore be a betterapproach to distinguishing cancer volumes from normal cellswhen they are partially mixed as in our case
Experimental attempts to obtain the optimal 1198791contrast
between cancerous and healthy tissue can be supportedby microscopic physiology studies on cancer cells As theWarburg effect states [32] cancer cells producemostly lactateat the end of their metabolic cycles while normal cellsmostly produce ATPs as an energy source The differentiatedby-products of the metabolic cycle may involve differentinfluences on the relaxation mechanism of water moleculesnear to those chemicals For instance the relaxation rate ofwatermolecules in protein is affected by themolecular weightof the protein [33] Justification for applying a low magneticfield rather than a high field will finally come only after abetter understanding of the relaxation process between by-products of metabolites and water molecules in cells To thisend further physiology-orientated studies at low magneticfields should be conducted in the future
The limitations in the interpretation of our results existin the present work First the measured 119879
1values of several
tissues in Table 1 exhibit rather large variation in the esti-mated errors Currently we suspect that the long averagingtime typically 20min for a single 119879
1measurement may
cause thermally induced problems such as slight variance ordegradation of 119861
119901due to the heated coil That is reasonable if
the circulation of LN2 is not smooth in an instant due to thelocal heating of 119861
119901coil although there is no heat dissipation
at the sample position by using a specially designed 119861119901
coil system Therefore it is definitely required to reduce theaveraging time in order to minimize the unwanted thermaleffects during the measurements Second the comparisonbetween benign and malignant tissues is missing In thiswork malignant tissues are compared only with normaltissues Third the fifteen pairs of tissues investigated inthis work may not seem sufficient for a reliable statisticalanalysis Particularly the result obtained from the mucinoustissue should be confirmed by investigating more tissuesalthough mucinous type is rare Finally the collected tissueswere frozen and preserved in LN2 dewar before the 119879
1
measurementsThe freezing and thawing processmay cause adamage in the tissues In future study amethod of preformingrelaxation measurement before freezing is needed to bedevised
In terms of SNR we believe two points can be improvedwhich will render single voxel relaxometry of breast cancerfeasible First the environmental noise of our system includ-ing SQUID was measured to be 23 fTHz12 and this valueis relatively higher than the other grouprsquos noise environmentbelow 1 fTHz12 [13 25] Second the SNR will increase ifthe 119861119901strength is increased by optimizing the coil system
Currently it was limited to a maximum of 60mT due to heatdissipation from the coil
5 Conclusions
In this study we performed 1198791relaxation measurements
of ex-vivo breast cancer tissue samples in order to verifythe feasibility of confirmative MR imaging of breast cancer
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
2 BioMed Research International
mediums the human brain have shown that the largest1198791contrast is accessible at low magnetic fields that is at
less than 300 120583T [9] An agarose gel solution has showndrastically enhanced 119879
1contrasts at magnetic fields of less
than 100 120583T [10 11]The1198791-weighted imaging of the tissue of a
rat tumor was obtained at approximately 100 120583T [12] Recent1198791measurements of prostate tissue at 132120583T have revealed
that the 1198791contrast has a linear dependence on the ratio
of the cancer volume in the tissue samples [13] Inspired bythese previous studies we focus on ex-vivo relaxometry ofbreast cancer tissue samples at low magnetic field conditionshere Since the mammary gland is functionally and quali-tatively similar to prostate tissue an enhanced 119879
1contrast
of breast cancer tissue can be expected at such low fieldsOur experimental procedure is conducted on a homebuiltultralow field (ULF)-nuclear magnetic resonance (NMR)setup [14ndash19] using superconducting quantum interferencedevice (SQUID) to sense weakNMR signals owing to the lowstatic magnetic field
SQUID-based microtesla NMR has paved the way fornew biomedical measurements under low magnetic fieldconditions for example chemical analysis from a 119869-couplingspectrum [20] simultaneous proton density imaging withmagnetoencephalography [21 22] direct neutral currentimaging [23ndash25] brainwave magnetic resonance [18] andheartmagnetic resonance [26] All of thesemeasurements aredifficult to achieve with conventional high-field NMRMRIsThe present studyrsquos aim to achieve confirmative MR imagingof cancer tissues in-vivo and contrast agent free will certainlybe one of the most fruitful applications of low field MRI
In this study we measured the 1198791contrasts of breast
cancer tissues at three different low fields 37 62 and122120583T The 119879
1trend variation may give us a relaxometric
fingerprint to discriminate between breast cancer and healthytissue either by estimating an optimal magnetic field or bycomparing the slope of the 119879
1variation along the different
external magnetic fields Since the tissue volumes were small(approximately 1 cm3) relatively low signal-to-noise ratios(SNR) were obtained Hence we used a signal reconstructionmethod based on Bayesian statistics for noise removal andreliably separated signal amplitudes from noise Accordingto the model for relaxation in a multiphase system [27]provided that the exchange rate is faster than the relaxationrate (1119879
1) the relaxation rate has a linear dependence on the
volume ratio and the coefficient of the linearity provides the1198791contrast between themedia From the same analysis of our
data the 1198791contrasts at three magnetic fields failed to exhibit
ameaningful relevance to the volume ratio of breast cancer intissue One result worth noting is that mucinous carcinomatissue can be distinguished clearly from other tissue becauseof its high water density and long 119879
1relaxation time Possible
implications of our results and further studywill be discussed
2 Methods
21 Sample Treatment In this study fifteen pairs of breasttissue specimens were investigated Each pair consists of
nominal tumor and normal tissues All the tissue sam-ples were obtained from total mastectomy specimens inAjou University Hospital (Suwon Korea) The tumor tissuesamples were found to include various cancer types teninvasive ductal carcinomas one ductal carcinoma in situone mixed ductal and lobular carcinoma and one mucinouscarcinoma Apart from the mucinous case discriminationbetween cancer types was not the concern of this study Thisresearch was approved by the Institutional Review Board ofAjou University Hospital
After gross examination of each mastectomy specimenrepresentative tumor and normal tissue samples with vol-umes of approximately 1 cm3 were prepared and then imme-diately inserted into a liquid nitrogen (LN2) tank where theywere stored until the NMR experiments were conductedfollowing Huang et al [8] Prior to the 119879
1measurements
the tissues were thawed in the ambient atmosphere andtemperature To avoid dehydration after thawing they werewrapped in cling film
The specimens were formalin-fixed after the NMR mea-surements were complete The paired tumor and normaltissue samples were routinely processed and stained withhematoxylin-eosin for histologic examination From eachtissue a single slide was prepared via sectioning and exam-ined with a microscope The proportion ratios of cancer cellswere measured by mapping the slides on a square grid inmillimeter
22 Experimental Setup Figure 1 shows the experimentalsetup for the microtesla NMR experiment All equipmentwas mounted inside a magnetically shielded room (MSR)which was specially designed for ULF-NMRMRI [18] A dc-SQUID (SupraconAG)was used as aNMR signal detector forthe ULF-NMR system The dc-SQUID sensor was shieldedwith two different cans The inner shield is a superconduct-ing Nb cast can with 999 purity The outer shield is asuperconducting lead can with 995 purity The pickup coilis a second-order gradiometer which is made of a 125120583mNb wire with a 50mm baseline and a diameter of 65mmThe dc-SQUID with pickup coil was additionally wrappedwith an aluminum-coated Mylar film to prevent the SQUIDsystem from being influenced by ambient RF noise Theenvironmental noise of our system inside the MSRmeasuredby the dc-SQUID sensor with second-order gradiometricpickup coil is about 23 fTHz12 at 100Hz
Since magnetization in the microtesla region is insuffi-cient to produce the NMR signal in the ULF-NMR a strongmagnetic field (119861
119901) for prepolarizing the nuclear magnetiza-
tion must be applied prior to NMR signal acquisitionThe 119861119901
strength is in tens of mT The free precession of the nuclearmagnetization is then recorded under a measurement field(119861119898) in the 120583T region A double Helmholtz coil [28] was
used to produce a high degree of homogeneity in the 119861119898
The119861119898strength at the sample space was about 493120583T which
corresponds to a proton NMR frequency of about 210HzProducing a high 119861
119901leads to the heating of the 119861
119901coil
due to the injection of a high current For the measurementof the ex-vivo tissues the temperature increase in the tis-sue caused by heat transfer from the 119861
119901coil may induce
BioMed Research International 3
SQUID dewar
Reservoir
N2 gas
LN2
1st Bp2nd Bp
Bm
x
y
z
dewar
Figure 1 Schematic diagram of the experimental apparatus for themicrotesla NMR experimentThe axes of the 1st and 2nd 119861
119901coils are
parallel However these are perpendicular to the axis of the 119861119898coil
The 1st 119861119901coil is a liquid nitrogen (LN2)-cooled pancake-type coil
The LN2 flowed continuously from the reservoir to the 1st 119861119901dewar
through two tubes at the bottom and then the nitrogen gas flowedback to the reservoir through the upper tube
undesired effects such as tissue degeneration and variationin 1198791relaxation time It is therefore important to maintain
a constant tissue temperature during the measurements The119861119901coil system corresponding to the 1st 119861
119901dewar in Figure 1
was designed for this purposeFor effective generation of the 119861
119901at the sample region
which is beneath the base of the SQUID dewar we adopted apancake-type geometry for the 119861
119901coil A 700-turn copper-
wire-wound pancake coil which has an outer diameter of132mm and a length of 16mm generates a magnetic field of60mT at the region of the sample with a current injection of20A
The 119861119901coil was mounted in a dewar which is the green
cylindrical component labeled ldquo1st 119861119901dewarrdquo in Figure 1 for
effective cooling Liquid nitrogen (ldquoLN2rdquo in Figure 1) flowscontinuously from the reservoir to cool the 119861
119901coil Nitrogen
gas produced by the resultant boiled liquid then flows backout to the reservoir The multiple layers of glass fiber inthe covering plate of the dewar prevent the refrigerationof the tissue A 3D-printed 4-mm thick thermal shieldplate having a meandering air path was additionally placedbetween the sample and the covering plate of the 119861
119901dewar
in order to keep the tissue temperature constant against therelatively cold surface of the dewarTheN
2gas was channeled
continuously through the thermal shield plate resulting in atemperature variation of plusmn03∘C for thirty minutes on the topsurface of the plate
A Helmholtz coil was used for the 2nd 119861119901 The field
strengths were varied between 37 62 and 122 120583T whichwere measured at the sample space The 119861
119898and the 2nd 119861
119901
coils were controlled by solid state relays (SSR) Howevera large amount of power is necessary to generate 1st 119861
119901of
approximately 60mT which should be switched off after a
Signal
1st Bp
2nd Bp
Bm
75ms 10ms
tB119901 tdelay tmeasurement trepetition
Figure 2 Illustration of pulse sequence for1198791experimentsThe data
set of 1198791is obtained by changing the duration of 119905delay
few milliseconds Therefore the 1st 119861119901coil was controlled
by a specially constructed current-driving circuit [17] whichis basically composed of a capacitor bank and an insulatedgate bipolar transistor [29]The SSRs and the current-drivingcircuit were remotely switched by a programmable pulse-generating board connected by optical fibers to protect theULF-NMR system from external electronic noise
Figure 2 shows the pulse sequence for the 1198791experiment
Initially the 1st 119861119901is applied to form net magnetization
in one direction After the 1st 119861119901is turned off the spins
begin to relax while the 2nd 119861119901is turned on After the
time 119905delay has elapsed the spins produce a free precessiondecay (FPD) signal which is measured by the SQUID sensorSince the degree of relaxation differs in accordance with theduration of 119905delay a 1198791 curve is obtained by varying 119905delay Thefollowing experimental parameters were used in this study(see Figure 2) 119905
119861119901= 1 s 119905measurement = 1 s and 119905repetition = 7 s
The averaged FPD signals were measured at each time 119905delayThe number of averages was mainly 20sim40 depending on thesample We measured mainly 4 points with different 119905delayvalues for each sample in order to generate the 119879
1curve
23 Analysis MethodmdashBayesian Analysis ANMR signal canbe expressed as the summation of 119873 exponentially decayingsinusoids and Gaussian noise [30]
119889119896=
119873
sum
119895=1
[(119860119895119890119894120601119895) 119890minus119896Δ119905120591119895
1198902120587119894119896Δ119905119891119895
] + 120576119896
119896 = 0 119872 minus 1
(1)
where 119860119895is the signal amplitude 120601
119895is the phase 120591
119895is the
decay time 119891119895is the frequency of the 119895th sinusoid119872 is the
number of data points and 120576119896is Gaussian random noise with
zero mean and standard deviation 120590 Since this informationabout the NMR signal is intensively located in the beginningpart of the FPD signal it is therefore important to record theinitial part of that signal
If the SNR is low even the initial signal is already hiddenby noise as was the case for the tissue samples investigatedin this study Therefore we adopted the Bayesian analysismethod [30] which is suitable for the analysis of small signalsThe primary assumption is that the noise obtained during thedata acquisition has a Gaussian distribution After adopting
4 BioMed Research International
the Bayesian rule the probability becomes a function of thedecay time and the resonance frequency Consider
119875 (d | 120591 f 120590119873)
=
(21205871205902)119873minus119872
1205821sdot sdot sdot 120582119873
times exp[[
[
minussum119872minus1
119896=0
10038161003816100381610038161198891198961003816100381610038161003816
2
+ sum119873
119895=1120582119895
100381610038161003816100381610038161003816(ddaggerUE)
lowast
119895120582119895
100381610038161003816100381610038161003816
2
21205902
]]
]
(2)
where U = exp(minus119896Δ119905120591119895) exp(2120587119894119896Δ119905119891
119895) and E and 120582 are
the matrices of the eigenvectors and eigenvalues of UdaggerUrespectively The equation of probability in [30] needs to beadjusted as illustrated in (2) Therefore the solution is to findthe maximum probability by varying two variables the decaytime and the resonance frequency When the probability isat a maximum the signal intensity and the phase are alsodetermined as
119860119895119890119894120601119895=
[E (ddaggerUE)lowast
]119895
120582119895
(3)
Since the resonance frequency can be determined by theFFT of the relatively larger signal or the measurement ofthe 119861119898strength the variable of the resonance frequency is
additionally reducedTherefore more precise information onthe signal intensity could be obtained from the FPD signalDuring the analysis of the acquired data nondecaying termssuch as the higher harmonics of the AC line noise andorthe additional noise were also considered by adding theseterms to the equations whereas only one decaying termcorresponding to the FPD signal was used
3 Results
To estimate 1198791relaxation times the variation in the FPD
signal amplitudes was measured as a function of the timeduration 119905delay Each data set was fitted using the followingequation
119891 (119905) = 120572 exp(minus119905delay
1198791
) (4)
where 120572 is the amplitude and 1198791is the longitudinal relaxation
timeThe additional constant to represent the base-offset wasnot included because the Bayesian analysis described abovediscriminates NMR signals from noisy backgrounds well(see Figure 3 andor Supplementary Information in Supple-mentaryMaterial available online at httpdxdoiorg1011552014385428)
Table 1 summarizes the pathologic characteristics frac-tion of cancerous cells in each tumor and normal tissue (therest of the tissue is filled with normal cells) and themeasured1198791values at three magnetic fields of the thirteen pairs of
breast tissue specimens which exhibited FPD signals The
Original
00 01 02 03
Time (s)
1
0
minus1
FPD
sign
al (a
u)
(a)
With noise
00 01 02 03
Time (s)
3
0
minus3
FPD
sign
al (a
u)
(b)
Reconstructed
00 01 02 03
Time (s)
1
0
minus1FP
D si
gnal
(au
)
(c)
Figure 3 Simulation of Bayesian analysis (a) the simulated FPDsignal (b) the mixed signal with Gaussian random noise and (c)the reconstructed signal from the noisy signal plotted in (b) Thesimulated FPD signal was built on assumptions of 119860
119895= 1 120601
119895= 0
120591119895= 008 and 119891
119895= 211 in (1) The nondecaying signals were also
added as electronic noise which had the same amplitude of 002and frequencies of 180 and 240Hz The simulated FPD signal wasmixed with Gaussian random noise which had a standard deviationof 08 After Bayesian analysis the reconstructed signal had thedecaying components 119860
119895= 1031 120601
119895= 0011 120591
119895= 0066 and
119891119895= 211 The nondecaying components of the reconstructed signal
were composed of the amplitudes and phases of 119860180
= 0026119860240= 0027 and 120601
180= minus0857 120601
240= 0314 respectively
examples of 1198791fitted curves are shown in the Supplementary
Information section In addition two 1198791values could not
be estimated using the fitting equation because the numberof data points obtained by the FPD signal were insufficientIn those cases the baseline of the FPD signal fluctuates andis discontinued Most tumor specimens were composed ofcancer cells surrounding desmoplastic stroma and somenormal epithelial cells For example the invasive ductalcarcinoma (Specimen number S13-10) plotted in Figure 4(a)shows irregularly infiltrating epithelia cell clusters and awell-organized normal breast lobule which are denoted bythe arrows on the right- and left-hand sides of the picturerespectively The mucinous carcinoma specimen howeverhad only cancer cells and mucin with no normal tissueThemucinous carcinoma sample (Specimen number S13-09)plotted in Figure 4(b) showed a few clusters of malignantepithelial cells in the background of the mucin pool whereasthe normal breast tissue was composed of fibrotic stroma anda few epithelial cells (see Figure 4(c)) Among the four cases
BioMed Research International 5
Table1Analyzedandmeasuredparameters
Specim
ennu
mber
Type
Cancer
1198791at122120583
T(m
s)1198791at62120583T(m
s)1198791at37120583T(m
s)Tu
mor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
S13-01
Ductal
900
604plusmn457
482plusmn264
841plusmn219
914plusmn422
1095plusmn429
69plusmn3
S13-02
aDuctal
250
1611plusmn532
813plusmn247
535plusmn108
941plusmn332
79plusmn352
564plusmn193
S13-03
Ductal
750
687plusmn99
74plusmn385
634plusmn29
623plusmn205
668plusmn137
977plusmn184
S13-04
aLo
bular
600
81plusmn201
546plusmn11
618plusmn5
608plusmn5
595plusmn89
66plusmn84
S13-05
aDuctal+
lobu
lar
605
1001plusmn
74mdash
121plusmn
533
866plusmn5
685plusmn104
105plusmn124
S13-06
Ductal
700
722plusmn228
673plusmn75
636plusmn138
676plusmn83
794plusmn52
634plusmn53
S13-07
Ductal
750
792plusmn58
1013plusmn112
736plusmn48
744plusmn123
682plusmn48
742plusmn64
S13-08
Ductal
900
1692plusmn147
833plusmn156
819plusmn283
829plusmn93
641plusmn184
612plusmn128
S13-09
Mucinou
s100
03477plusmn249
775plusmn269
3586plusmn258
847plusmn134
mdash46
3plusmn56
S13-10
Ductal
750
67plusmn3
635plusmn97
689plusmn27
844plusmn19
656plusmn24
659plusmn56
S13-11
aDuctal
900
861plusmn79
984plusmn298
84plusmn61
603plusmn44
3765plusmn13
876plusmn278
S13-12
Ductal
700
115plusmn117
1006plusmn85
912plusmn33
812plusmn27
498plusmn71
255plusmn78
S13-13
Ductal
900
655plusmn03
758plusmn212
522plusmn154
773plusmn81
64plusmn31
816plusmn145
a Thep
reop
erativec
hemotherapy
was
perfo
rmed
6 BioMed Research International
(a) (b)
(c)
Figure 4 Photomicrographs of (a) the invasive ductal carcinoma sample (Specimen number S13-10) (b) the mucinous carcinoma sample(Specimen number S13-09) and (c) normal breast tissue (Specimen number S13-09) The arrows on the right- and left-hand sides of thepicture in (a) represent irregularly infiltrating epithelia cell clusters and a well-organized normal breast lobule respectively Each specimenwas stained with hematoxylin-eosin and observed at times40 magnification
receiving preoperative chemotherapy one case (Specimennumber S13-02) responded to treatment and the majority ofthe tumor tissue was replaced with fibrotic stroma
Since the tumor tissues contain partial cancers the 1198791
values estimated from the fitting do not yield meaningfulinformation about the 119879
1contrast unless properly analyzed
For correct estimation of the 1198791contrast between cancer and
normal tissue the cancer volume ratio must be considered aspart of the analysis The model devised by Zimmerman andBrittin [27] describes the relaxation phenomenon in a systemof multiple phases According to this model the exchangerate between the twomedia with different119879
1relaxation times
strongly influences the 1198791relaxation process We presume
that the exchange rate of the water molecules betweennormal and cancer cells in a tumor tissue is reasonablyrapid compared with the relaxation rate 1119879
1 Then a single
exponential decay with the average relaxation rate 11198791119879=
1205881198621198791119862+ 1205881198731198791119873 will be obtained in which 119879
1119862and 119879
1119873
are the 1198791relaxation times of the cancer and normal cells
respectively The 120588119862and 120588
119873are volume ratios of the cancer
and normal cells in the tissueThe ratio between1198791119873
and1198791119879
that is 11987911198731198791119879 of each tumor tissue should show a linear
dependence on the 120588119862and the coefficient is the 119879
1contrast
120575 between the cancer and normal cells as
1198791119873
1198791119879
= 1 + 120588119862(1198791119873
1198791119862
minus 1) = 1 + 120588119862120575 (5)
Figure 5 shows the 1198791ratios between the tumor and the
healthy tissue (11987911198731198791119879) as a function of the relative cancer
volume 120588119862 The (red) closed circles (blue) open circles and
(green) closed squares represent the data obtained at 122 62and 37 120583T respectively Apparently the119879
1relaxation time has
a weak dependency on 120588119862 and the 119879
1contrast 120575 exhibits a
weak variance as the magnetic field increases The estimated120575 is 33 times 10minus4 at 37 120583T 63 times 10minus4 at 62120583T and minus94 times 10minus4 at122 120583T which do not appear to bemeaningful compared withstandard deviations of the data Possible reasons for this willbe discussed below
From the mucinous carcinoma tissue (Specimen numberS13-09) a clearly distinguishable NMR signal was measuredas shown in Figure 6 The NMR spectrum has a full-widthhalf-maximum (FWHM) of 17Hz which is about onequarter of the FWHMs of the other tissue which are typically7Hz As a result 120575 is higher than 4 As in the case of high-field MRI [31] mucinous carcinoma tissue can be identifiedunambiguously in low field MRI via 119879
1or 1198792weighted
imaging methods
4 Discussion
In the multiphases relaxation model [27] (5) is based onthe assumption that water molecules in cancer and normalcells can be exchanged rapidly compared with the relaxation
BioMed Research International 7
20 30 40 50 60 70 80 90
Ratio of cancer ()
2
1
0
T1
NT
1T
Figure 5 The 1198791ratio between the tumor (T) and the normal (N)
tissue as a function of the cancer ratio Since the normal tissue inspecimen S13-05 was 5 cancerous the surface ratio of the cancerof this specimen was denoted as 55 The (red) closed circles(blue) open circles and (green) closed squares represent the dataobtained at 122 62 and 37 120583T respectively The error bars indicatethe standard errors The lines are linear fits to data obtained at eachfield
190 200 210 220 230000
005
010
015
020
025
Nor
mal
ized
FFT
(au
)
Frequency (Hz)
FWHM 7Hz
FWHM 17Hz
Figure 6 The normalized spectra for the tumor tissues of S13-09(red) and S13-10 (green) obtained by the experimental conditions of119905delay = 10ms and 2nd 119861
119901= 62 120583T The spectra were normalized
with each spectrum area
rates of water molecules belonging to separate phases in thesystem If this assumption is not valid the 119879
1relaxation of
tissue should exhibitmultiple exponentialsThe1198791values dis-
played in Table 1 are mostly lt100msTherefore the exchangerate between water molecules in cancer and normal cellsshould be at least 10 sminus1 in order for (5) to be applied Suchan exchange rate might not be a realistic value We speculatethat cancer cells are still isolated from normal cells for longerperiods than the time scale of 100ms In this case multipleexponential fitting of the decay curve should be required tocorrectly analyze tissue samples having a cancer ratio in therange of 20sim80The relaxometry of a single (or finite) voxel
in the MR imaging of tumor tissue will therefore be a betterapproach to distinguishing cancer volumes from normal cellswhen they are partially mixed as in our case
Experimental attempts to obtain the optimal 1198791contrast
between cancerous and healthy tissue can be supportedby microscopic physiology studies on cancer cells As theWarburg effect states [32] cancer cells producemostly lactateat the end of their metabolic cycles while normal cellsmostly produce ATPs as an energy source The differentiatedby-products of the metabolic cycle may involve differentinfluences on the relaxation mechanism of water moleculesnear to those chemicals For instance the relaxation rate ofwatermolecules in protein is affected by themolecular weightof the protein [33] Justification for applying a low magneticfield rather than a high field will finally come only after abetter understanding of the relaxation process between by-products of metabolites and water molecules in cells To thisend further physiology-orientated studies at low magneticfields should be conducted in the future
The limitations in the interpretation of our results existin the present work First the measured 119879
1values of several
tissues in Table 1 exhibit rather large variation in the esti-mated errors Currently we suspect that the long averagingtime typically 20min for a single 119879
1measurement may
cause thermally induced problems such as slight variance ordegradation of 119861
119901due to the heated coil That is reasonable if
the circulation of LN2 is not smooth in an instant due to thelocal heating of 119861
119901coil although there is no heat dissipation
at the sample position by using a specially designed 119861119901
coil system Therefore it is definitely required to reduce theaveraging time in order to minimize the unwanted thermaleffects during the measurements Second the comparisonbetween benign and malignant tissues is missing In thiswork malignant tissues are compared only with normaltissues Third the fifteen pairs of tissues investigated inthis work may not seem sufficient for a reliable statisticalanalysis Particularly the result obtained from the mucinoustissue should be confirmed by investigating more tissuesalthough mucinous type is rare Finally the collected tissueswere frozen and preserved in LN2 dewar before the 119879
1
measurementsThe freezing and thawing processmay cause adamage in the tissues In future study amethod of preformingrelaxation measurement before freezing is needed to bedevised
In terms of SNR we believe two points can be improvedwhich will render single voxel relaxometry of breast cancerfeasible First the environmental noise of our system includ-ing SQUID was measured to be 23 fTHz12 and this valueis relatively higher than the other grouprsquos noise environmentbelow 1 fTHz12 [13 25] Second the SNR will increase ifthe 119861119901strength is increased by optimizing the coil system
Currently it was limited to a maximum of 60mT due to heatdissipation from the coil
5 Conclusions
In this study we performed 1198791relaxation measurements
of ex-vivo breast cancer tissue samples in order to verifythe feasibility of confirmative MR imaging of breast cancer
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
BioMed Research International 3
SQUID dewar
Reservoir
N2 gas
LN2
1st Bp2nd Bp
Bm
x
y
z
dewar
Figure 1 Schematic diagram of the experimental apparatus for themicrotesla NMR experimentThe axes of the 1st and 2nd 119861
119901coils are
parallel However these are perpendicular to the axis of the 119861119898coil
The 1st 119861119901coil is a liquid nitrogen (LN2)-cooled pancake-type coil
The LN2 flowed continuously from the reservoir to the 1st 119861119901dewar
through two tubes at the bottom and then the nitrogen gas flowedback to the reservoir through the upper tube
undesired effects such as tissue degeneration and variationin 1198791relaxation time It is therefore important to maintain
a constant tissue temperature during the measurements The119861119901coil system corresponding to the 1st 119861
119901dewar in Figure 1
was designed for this purposeFor effective generation of the 119861
119901at the sample region
which is beneath the base of the SQUID dewar we adopted apancake-type geometry for the 119861
119901coil A 700-turn copper-
wire-wound pancake coil which has an outer diameter of132mm and a length of 16mm generates a magnetic field of60mT at the region of the sample with a current injection of20A
The 119861119901coil was mounted in a dewar which is the green
cylindrical component labeled ldquo1st 119861119901dewarrdquo in Figure 1 for
effective cooling Liquid nitrogen (ldquoLN2rdquo in Figure 1) flowscontinuously from the reservoir to cool the 119861
119901coil Nitrogen
gas produced by the resultant boiled liquid then flows backout to the reservoir The multiple layers of glass fiber inthe covering plate of the dewar prevent the refrigerationof the tissue A 3D-printed 4-mm thick thermal shieldplate having a meandering air path was additionally placedbetween the sample and the covering plate of the 119861
119901dewar
in order to keep the tissue temperature constant against therelatively cold surface of the dewarTheN
2gas was channeled
continuously through the thermal shield plate resulting in atemperature variation of plusmn03∘C for thirty minutes on the topsurface of the plate
A Helmholtz coil was used for the 2nd 119861119901 The field
strengths were varied between 37 62 and 122 120583T whichwere measured at the sample space The 119861
119898and the 2nd 119861
119901
coils were controlled by solid state relays (SSR) Howevera large amount of power is necessary to generate 1st 119861
119901of
approximately 60mT which should be switched off after a
Signal
1st Bp
2nd Bp
Bm
75ms 10ms
tB119901 tdelay tmeasurement trepetition
Figure 2 Illustration of pulse sequence for1198791experimentsThe data
set of 1198791is obtained by changing the duration of 119905delay
few milliseconds Therefore the 1st 119861119901coil was controlled
by a specially constructed current-driving circuit [17] whichis basically composed of a capacitor bank and an insulatedgate bipolar transistor [29]The SSRs and the current-drivingcircuit were remotely switched by a programmable pulse-generating board connected by optical fibers to protect theULF-NMR system from external electronic noise
Figure 2 shows the pulse sequence for the 1198791experiment
Initially the 1st 119861119901is applied to form net magnetization
in one direction After the 1st 119861119901is turned off the spins
begin to relax while the 2nd 119861119901is turned on After the
time 119905delay has elapsed the spins produce a free precessiondecay (FPD) signal which is measured by the SQUID sensorSince the degree of relaxation differs in accordance with theduration of 119905delay a 1198791 curve is obtained by varying 119905delay Thefollowing experimental parameters were used in this study(see Figure 2) 119905
119861119901= 1 s 119905measurement = 1 s and 119905repetition = 7 s
The averaged FPD signals were measured at each time 119905delayThe number of averages was mainly 20sim40 depending on thesample We measured mainly 4 points with different 119905delayvalues for each sample in order to generate the 119879
1curve
23 Analysis MethodmdashBayesian Analysis ANMR signal canbe expressed as the summation of 119873 exponentially decayingsinusoids and Gaussian noise [30]
119889119896=
119873
sum
119895=1
[(119860119895119890119894120601119895) 119890minus119896Δ119905120591119895
1198902120587119894119896Δ119905119891119895
] + 120576119896
119896 = 0 119872 minus 1
(1)
where 119860119895is the signal amplitude 120601
119895is the phase 120591
119895is the
decay time 119891119895is the frequency of the 119895th sinusoid119872 is the
number of data points and 120576119896is Gaussian random noise with
zero mean and standard deviation 120590 Since this informationabout the NMR signal is intensively located in the beginningpart of the FPD signal it is therefore important to record theinitial part of that signal
If the SNR is low even the initial signal is already hiddenby noise as was the case for the tissue samples investigatedin this study Therefore we adopted the Bayesian analysismethod [30] which is suitable for the analysis of small signalsThe primary assumption is that the noise obtained during thedata acquisition has a Gaussian distribution After adopting
4 BioMed Research International
the Bayesian rule the probability becomes a function of thedecay time and the resonance frequency Consider
119875 (d | 120591 f 120590119873)
=
(21205871205902)119873minus119872
1205821sdot sdot sdot 120582119873
times exp[[
[
minussum119872minus1
119896=0
10038161003816100381610038161198891198961003816100381610038161003816
2
+ sum119873
119895=1120582119895
100381610038161003816100381610038161003816(ddaggerUE)
lowast
119895120582119895
100381610038161003816100381610038161003816
2
21205902
]]
]
(2)
where U = exp(minus119896Δ119905120591119895) exp(2120587119894119896Δ119905119891
119895) and E and 120582 are
the matrices of the eigenvectors and eigenvalues of UdaggerUrespectively The equation of probability in [30] needs to beadjusted as illustrated in (2) Therefore the solution is to findthe maximum probability by varying two variables the decaytime and the resonance frequency When the probability isat a maximum the signal intensity and the phase are alsodetermined as
119860119895119890119894120601119895=
[E (ddaggerUE)lowast
]119895
120582119895
(3)
Since the resonance frequency can be determined by theFFT of the relatively larger signal or the measurement ofthe 119861119898strength the variable of the resonance frequency is
additionally reducedTherefore more precise information onthe signal intensity could be obtained from the FPD signalDuring the analysis of the acquired data nondecaying termssuch as the higher harmonics of the AC line noise andorthe additional noise were also considered by adding theseterms to the equations whereas only one decaying termcorresponding to the FPD signal was used
3 Results
To estimate 1198791relaxation times the variation in the FPD
signal amplitudes was measured as a function of the timeduration 119905delay Each data set was fitted using the followingequation
119891 (119905) = 120572 exp(minus119905delay
1198791
) (4)
where 120572 is the amplitude and 1198791is the longitudinal relaxation
timeThe additional constant to represent the base-offset wasnot included because the Bayesian analysis described abovediscriminates NMR signals from noisy backgrounds well(see Figure 3 andor Supplementary Information in Supple-mentaryMaterial available online at httpdxdoiorg1011552014385428)
Table 1 summarizes the pathologic characteristics frac-tion of cancerous cells in each tumor and normal tissue (therest of the tissue is filled with normal cells) and themeasured1198791values at three magnetic fields of the thirteen pairs of
breast tissue specimens which exhibited FPD signals The
Original
00 01 02 03
Time (s)
1
0
minus1
FPD
sign
al (a
u)
(a)
With noise
00 01 02 03
Time (s)
3
0
minus3
FPD
sign
al (a
u)
(b)
Reconstructed
00 01 02 03
Time (s)
1
0
minus1FP
D si
gnal
(au
)
(c)
Figure 3 Simulation of Bayesian analysis (a) the simulated FPDsignal (b) the mixed signal with Gaussian random noise and (c)the reconstructed signal from the noisy signal plotted in (b) Thesimulated FPD signal was built on assumptions of 119860
119895= 1 120601
119895= 0
120591119895= 008 and 119891
119895= 211 in (1) The nondecaying signals were also
added as electronic noise which had the same amplitude of 002and frequencies of 180 and 240Hz The simulated FPD signal wasmixed with Gaussian random noise which had a standard deviationof 08 After Bayesian analysis the reconstructed signal had thedecaying components 119860
119895= 1031 120601
119895= 0011 120591
119895= 0066 and
119891119895= 211 The nondecaying components of the reconstructed signal
were composed of the amplitudes and phases of 119860180
= 0026119860240= 0027 and 120601
180= minus0857 120601
240= 0314 respectively
examples of 1198791fitted curves are shown in the Supplementary
Information section In addition two 1198791values could not
be estimated using the fitting equation because the numberof data points obtained by the FPD signal were insufficientIn those cases the baseline of the FPD signal fluctuates andis discontinued Most tumor specimens were composed ofcancer cells surrounding desmoplastic stroma and somenormal epithelial cells For example the invasive ductalcarcinoma (Specimen number S13-10) plotted in Figure 4(a)shows irregularly infiltrating epithelia cell clusters and awell-organized normal breast lobule which are denoted bythe arrows on the right- and left-hand sides of the picturerespectively The mucinous carcinoma specimen howeverhad only cancer cells and mucin with no normal tissueThemucinous carcinoma sample (Specimen number S13-09)plotted in Figure 4(b) showed a few clusters of malignantepithelial cells in the background of the mucin pool whereasthe normal breast tissue was composed of fibrotic stroma anda few epithelial cells (see Figure 4(c)) Among the four cases
BioMed Research International 5
Table1Analyzedandmeasuredparameters
Specim
ennu
mber
Type
Cancer
1198791at122120583
T(m
s)1198791at62120583T(m
s)1198791at37120583T(m
s)Tu
mor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
S13-01
Ductal
900
604plusmn457
482plusmn264
841plusmn219
914plusmn422
1095plusmn429
69plusmn3
S13-02
aDuctal
250
1611plusmn532
813plusmn247
535plusmn108
941plusmn332
79plusmn352
564plusmn193
S13-03
Ductal
750
687plusmn99
74plusmn385
634plusmn29
623plusmn205
668plusmn137
977plusmn184
S13-04
aLo
bular
600
81plusmn201
546plusmn11
618plusmn5
608plusmn5
595plusmn89
66plusmn84
S13-05
aDuctal+
lobu
lar
605
1001plusmn
74mdash
121plusmn
533
866plusmn5
685plusmn104
105plusmn124
S13-06
Ductal
700
722plusmn228
673plusmn75
636plusmn138
676plusmn83
794plusmn52
634plusmn53
S13-07
Ductal
750
792plusmn58
1013plusmn112
736plusmn48
744plusmn123
682plusmn48
742plusmn64
S13-08
Ductal
900
1692plusmn147
833plusmn156
819plusmn283
829plusmn93
641plusmn184
612plusmn128
S13-09
Mucinou
s100
03477plusmn249
775plusmn269
3586plusmn258
847plusmn134
mdash46
3plusmn56
S13-10
Ductal
750
67plusmn3
635plusmn97
689plusmn27
844plusmn19
656plusmn24
659plusmn56
S13-11
aDuctal
900
861plusmn79
984plusmn298
84plusmn61
603plusmn44
3765plusmn13
876plusmn278
S13-12
Ductal
700
115plusmn117
1006plusmn85
912plusmn33
812plusmn27
498plusmn71
255plusmn78
S13-13
Ductal
900
655plusmn03
758plusmn212
522plusmn154
773plusmn81
64plusmn31
816plusmn145
a Thep
reop
erativec
hemotherapy
was
perfo
rmed
6 BioMed Research International
(a) (b)
(c)
Figure 4 Photomicrographs of (a) the invasive ductal carcinoma sample (Specimen number S13-10) (b) the mucinous carcinoma sample(Specimen number S13-09) and (c) normal breast tissue (Specimen number S13-09) The arrows on the right- and left-hand sides of thepicture in (a) represent irregularly infiltrating epithelia cell clusters and a well-organized normal breast lobule respectively Each specimenwas stained with hematoxylin-eosin and observed at times40 magnification
receiving preoperative chemotherapy one case (Specimennumber S13-02) responded to treatment and the majority ofthe tumor tissue was replaced with fibrotic stroma
Since the tumor tissues contain partial cancers the 1198791
values estimated from the fitting do not yield meaningfulinformation about the 119879
1contrast unless properly analyzed
For correct estimation of the 1198791contrast between cancer and
normal tissue the cancer volume ratio must be considered aspart of the analysis The model devised by Zimmerman andBrittin [27] describes the relaxation phenomenon in a systemof multiple phases According to this model the exchangerate between the twomedia with different119879
1relaxation times
strongly influences the 1198791relaxation process We presume
that the exchange rate of the water molecules betweennormal and cancer cells in a tumor tissue is reasonablyrapid compared with the relaxation rate 1119879
1 Then a single
exponential decay with the average relaxation rate 11198791119879=
1205881198621198791119862+ 1205881198731198791119873 will be obtained in which 119879
1119862and 119879
1119873
are the 1198791relaxation times of the cancer and normal cells
respectively The 120588119862and 120588
119873are volume ratios of the cancer
and normal cells in the tissueThe ratio between1198791119873
and1198791119879
that is 11987911198731198791119879 of each tumor tissue should show a linear
dependence on the 120588119862and the coefficient is the 119879
1contrast
120575 between the cancer and normal cells as
1198791119873
1198791119879
= 1 + 120588119862(1198791119873
1198791119862
minus 1) = 1 + 120588119862120575 (5)
Figure 5 shows the 1198791ratios between the tumor and the
healthy tissue (11987911198731198791119879) as a function of the relative cancer
volume 120588119862 The (red) closed circles (blue) open circles and
(green) closed squares represent the data obtained at 122 62and 37 120583T respectively Apparently the119879
1relaxation time has
a weak dependency on 120588119862 and the 119879
1contrast 120575 exhibits a
weak variance as the magnetic field increases The estimated120575 is 33 times 10minus4 at 37 120583T 63 times 10minus4 at 62120583T and minus94 times 10minus4 at122 120583T which do not appear to bemeaningful compared withstandard deviations of the data Possible reasons for this willbe discussed below
From the mucinous carcinoma tissue (Specimen numberS13-09) a clearly distinguishable NMR signal was measuredas shown in Figure 6 The NMR spectrum has a full-widthhalf-maximum (FWHM) of 17Hz which is about onequarter of the FWHMs of the other tissue which are typically7Hz As a result 120575 is higher than 4 As in the case of high-field MRI [31] mucinous carcinoma tissue can be identifiedunambiguously in low field MRI via 119879
1or 1198792weighted
imaging methods
4 Discussion
In the multiphases relaxation model [27] (5) is based onthe assumption that water molecules in cancer and normalcells can be exchanged rapidly compared with the relaxation
BioMed Research International 7
20 30 40 50 60 70 80 90
Ratio of cancer ()
2
1
0
T1
NT
1T
Figure 5 The 1198791ratio between the tumor (T) and the normal (N)
tissue as a function of the cancer ratio Since the normal tissue inspecimen S13-05 was 5 cancerous the surface ratio of the cancerof this specimen was denoted as 55 The (red) closed circles(blue) open circles and (green) closed squares represent the dataobtained at 122 62 and 37 120583T respectively The error bars indicatethe standard errors The lines are linear fits to data obtained at eachfield
190 200 210 220 230000
005
010
015
020
025
Nor
mal
ized
FFT
(au
)
Frequency (Hz)
FWHM 7Hz
FWHM 17Hz
Figure 6 The normalized spectra for the tumor tissues of S13-09(red) and S13-10 (green) obtained by the experimental conditions of119905delay = 10ms and 2nd 119861
119901= 62 120583T The spectra were normalized
with each spectrum area
rates of water molecules belonging to separate phases in thesystem If this assumption is not valid the 119879
1relaxation of
tissue should exhibitmultiple exponentialsThe1198791values dis-
played in Table 1 are mostly lt100msTherefore the exchangerate between water molecules in cancer and normal cellsshould be at least 10 sminus1 in order for (5) to be applied Suchan exchange rate might not be a realistic value We speculatethat cancer cells are still isolated from normal cells for longerperiods than the time scale of 100ms In this case multipleexponential fitting of the decay curve should be required tocorrectly analyze tissue samples having a cancer ratio in therange of 20sim80The relaxometry of a single (or finite) voxel
in the MR imaging of tumor tissue will therefore be a betterapproach to distinguishing cancer volumes from normal cellswhen they are partially mixed as in our case
Experimental attempts to obtain the optimal 1198791contrast
between cancerous and healthy tissue can be supportedby microscopic physiology studies on cancer cells As theWarburg effect states [32] cancer cells producemostly lactateat the end of their metabolic cycles while normal cellsmostly produce ATPs as an energy source The differentiatedby-products of the metabolic cycle may involve differentinfluences on the relaxation mechanism of water moleculesnear to those chemicals For instance the relaxation rate ofwatermolecules in protein is affected by themolecular weightof the protein [33] Justification for applying a low magneticfield rather than a high field will finally come only after abetter understanding of the relaxation process between by-products of metabolites and water molecules in cells To thisend further physiology-orientated studies at low magneticfields should be conducted in the future
The limitations in the interpretation of our results existin the present work First the measured 119879
1values of several
tissues in Table 1 exhibit rather large variation in the esti-mated errors Currently we suspect that the long averagingtime typically 20min for a single 119879
1measurement may
cause thermally induced problems such as slight variance ordegradation of 119861
119901due to the heated coil That is reasonable if
the circulation of LN2 is not smooth in an instant due to thelocal heating of 119861
119901coil although there is no heat dissipation
at the sample position by using a specially designed 119861119901
coil system Therefore it is definitely required to reduce theaveraging time in order to minimize the unwanted thermaleffects during the measurements Second the comparisonbetween benign and malignant tissues is missing In thiswork malignant tissues are compared only with normaltissues Third the fifteen pairs of tissues investigated inthis work may not seem sufficient for a reliable statisticalanalysis Particularly the result obtained from the mucinoustissue should be confirmed by investigating more tissuesalthough mucinous type is rare Finally the collected tissueswere frozen and preserved in LN2 dewar before the 119879
1
measurementsThe freezing and thawing processmay cause adamage in the tissues In future study amethod of preformingrelaxation measurement before freezing is needed to bedevised
In terms of SNR we believe two points can be improvedwhich will render single voxel relaxometry of breast cancerfeasible First the environmental noise of our system includ-ing SQUID was measured to be 23 fTHz12 and this valueis relatively higher than the other grouprsquos noise environmentbelow 1 fTHz12 [13 25] Second the SNR will increase ifthe 119861119901strength is increased by optimizing the coil system
Currently it was limited to a maximum of 60mT due to heatdissipation from the coil
5 Conclusions
In this study we performed 1198791relaxation measurements
of ex-vivo breast cancer tissue samples in order to verifythe feasibility of confirmative MR imaging of breast cancer
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
4 BioMed Research International
the Bayesian rule the probability becomes a function of thedecay time and the resonance frequency Consider
119875 (d | 120591 f 120590119873)
=
(21205871205902)119873minus119872
1205821sdot sdot sdot 120582119873
times exp[[
[
minussum119872minus1
119896=0
10038161003816100381610038161198891198961003816100381610038161003816
2
+ sum119873
119895=1120582119895
100381610038161003816100381610038161003816(ddaggerUE)
lowast
119895120582119895
100381610038161003816100381610038161003816
2
21205902
]]
]
(2)
where U = exp(minus119896Δ119905120591119895) exp(2120587119894119896Δ119905119891
119895) and E and 120582 are
the matrices of the eigenvectors and eigenvalues of UdaggerUrespectively The equation of probability in [30] needs to beadjusted as illustrated in (2) Therefore the solution is to findthe maximum probability by varying two variables the decaytime and the resonance frequency When the probability isat a maximum the signal intensity and the phase are alsodetermined as
119860119895119890119894120601119895=
[E (ddaggerUE)lowast
]119895
120582119895
(3)
Since the resonance frequency can be determined by theFFT of the relatively larger signal or the measurement ofthe 119861119898strength the variable of the resonance frequency is
additionally reducedTherefore more precise information onthe signal intensity could be obtained from the FPD signalDuring the analysis of the acquired data nondecaying termssuch as the higher harmonics of the AC line noise andorthe additional noise were also considered by adding theseterms to the equations whereas only one decaying termcorresponding to the FPD signal was used
3 Results
To estimate 1198791relaxation times the variation in the FPD
signal amplitudes was measured as a function of the timeduration 119905delay Each data set was fitted using the followingequation
119891 (119905) = 120572 exp(minus119905delay
1198791
) (4)
where 120572 is the amplitude and 1198791is the longitudinal relaxation
timeThe additional constant to represent the base-offset wasnot included because the Bayesian analysis described abovediscriminates NMR signals from noisy backgrounds well(see Figure 3 andor Supplementary Information in Supple-mentaryMaterial available online at httpdxdoiorg1011552014385428)
Table 1 summarizes the pathologic characteristics frac-tion of cancerous cells in each tumor and normal tissue (therest of the tissue is filled with normal cells) and themeasured1198791values at three magnetic fields of the thirteen pairs of
breast tissue specimens which exhibited FPD signals The
Original
00 01 02 03
Time (s)
1
0
minus1
FPD
sign
al (a
u)
(a)
With noise
00 01 02 03
Time (s)
3
0
minus3
FPD
sign
al (a
u)
(b)
Reconstructed
00 01 02 03
Time (s)
1
0
minus1FP
D si
gnal
(au
)
(c)
Figure 3 Simulation of Bayesian analysis (a) the simulated FPDsignal (b) the mixed signal with Gaussian random noise and (c)the reconstructed signal from the noisy signal plotted in (b) Thesimulated FPD signal was built on assumptions of 119860
119895= 1 120601
119895= 0
120591119895= 008 and 119891
119895= 211 in (1) The nondecaying signals were also
added as electronic noise which had the same amplitude of 002and frequencies of 180 and 240Hz The simulated FPD signal wasmixed with Gaussian random noise which had a standard deviationof 08 After Bayesian analysis the reconstructed signal had thedecaying components 119860
119895= 1031 120601
119895= 0011 120591
119895= 0066 and
119891119895= 211 The nondecaying components of the reconstructed signal
were composed of the amplitudes and phases of 119860180
= 0026119860240= 0027 and 120601
180= minus0857 120601
240= 0314 respectively
examples of 1198791fitted curves are shown in the Supplementary
Information section In addition two 1198791values could not
be estimated using the fitting equation because the numberof data points obtained by the FPD signal were insufficientIn those cases the baseline of the FPD signal fluctuates andis discontinued Most tumor specimens were composed ofcancer cells surrounding desmoplastic stroma and somenormal epithelial cells For example the invasive ductalcarcinoma (Specimen number S13-10) plotted in Figure 4(a)shows irregularly infiltrating epithelia cell clusters and awell-organized normal breast lobule which are denoted bythe arrows on the right- and left-hand sides of the picturerespectively The mucinous carcinoma specimen howeverhad only cancer cells and mucin with no normal tissueThemucinous carcinoma sample (Specimen number S13-09)plotted in Figure 4(b) showed a few clusters of malignantepithelial cells in the background of the mucin pool whereasthe normal breast tissue was composed of fibrotic stroma anda few epithelial cells (see Figure 4(c)) Among the four cases
BioMed Research International 5
Table1Analyzedandmeasuredparameters
Specim
ennu
mber
Type
Cancer
1198791at122120583
T(m
s)1198791at62120583T(m
s)1198791at37120583T(m
s)Tu
mor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
S13-01
Ductal
900
604plusmn457
482plusmn264
841plusmn219
914plusmn422
1095plusmn429
69plusmn3
S13-02
aDuctal
250
1611plusmn532
813plusmn247
535plusmn108
941plusmn332
79plusmn352
564plusmn193
S13-03
Ductal
750
687plusmn99
74plusmn385
634plusmn29
623plusmn205
668plusmn137
977plusmn184
S13-04
aLo
bular
600
81plusmn201
546plusmn11
618plusmn5
608plusmn5
595plusmn89
66plusmn84
S13-05
aDuctal+
lobu
lar
605
1001plusmn
74mdash
121plusmn
533
866plusmn5
685plusmn104
105plusmn124
S13-06
Ductal
700
722plusmn228
673plusmn75
636plusmn138
676plusmn83
794plusmn52
634plusmn53
S13-07
Ductal
750
792plusmn58
1013plusmn112
736plusmn48
744plusmn123
682plusmn48
742plusmn64
S13-08
Ductal
900
1692plusmn147
833plusmn156
819plusmn283
829plusmn93
641plusmn184
612plusmn128
S13-09
Mucinou
s100
03477plusmn249
775plusmn269
3586plusmn258
847plusmn134
mdash46
3plusmn56
S13-10
Ductal
750
67plusmn3
635plusmn97
689plusmn27
844plusmn19
656plusmn24
659plusmn56
S13-11
aDuctal
900
861plusmn79
984plusmn298
84plusmn61
603plusmn44
3765plusmn13
876plusmn278
S13-12
Ductal
700
115plusmn117
1006plusmn85
912plusmn33
812plusmn27
498plusmn71
255plusmn78
S13-13
Ductal
900
655plusmn03
758plusmn212
522plusmn154
773plusmn81
64plusmn31
816plusmn145
a Thep
reop
erativec
hemotherapy
was
perfo
rmed
6 BioMed Research International
(a) (b)
(c)
Figure 4 Photomicrographs of (a) the invasive ductal carcinoma sample (Specimen number S13-10) (b) the mucinous carcinoma sample(Specimen number S13-09) and (c) normal breast tissue (Specimen number S13-09) The arrows on the right- and left-hand sides of thepicture in (a) represent irregularly infiltrating epithelia cell clusters and a well-organized normal breast lobule respectively Each specimenwas stained with hematoxylin-eosin and observed at times40 magnification
receiving preoperative chemotherapy one case (Specimennumber S13-02) responded to treatment and the majority ofthe tumor tissue was replaced with fibrotic stroma
Since the tumor tissues contain partial cancers the 1198791
values estimated from the fitting do not yield meaningfulinformation about the 119879
1contrast unless properly analyzed
For correct estimation of the 1198791contrast between cancer and
normal tissue the cancer volume ratio must be considered aspart of the analysis The model devised by Zimmerman andBrittin [27] describes the relaxation phenomenon in a systemof multiple phases According to this model the exchangerate between the twomedia with different119879
1relaxation times
strongly influences the 1198791relaxation process We presume
that the exchange rate of the water molecules betweennormal and cancer cells in a tumor tissue is reasonablyrapid compared with the relaxation rate 1119879
1 Then a single
exponential decay with the average relaxation rate 11198791119879=
1205881198621198791119862+ 1205881198731198791119873 will be obtained in which 119879
1119862and 119879
1119873
are the 1198791relaxation times of the cancer and normal cells
respectively The 120588119862and 120588
119873are volume ratios of the cancer
and normal cells in the tissueThe ratio between1198791119873
and1198791119879
that is 11987911198731198791119879 of each tumor tissue should show a linear
dependence on the 120588119862and the coefficient is the 119879
1contrast
120575 between the cancer and normal cells as
1198791119873
1198791119879
= 1 + 120588119862(1198791119873
1198791119862
minus 1) = 1 + 120588119862120575 (5)
Figure 5 shows the 1198791ratios between the tumor and the
healthy tissue (11987911198731198791119879) as a function of the relative cancer
volume 120588119862 The (red) closed circles (blue) open circles and
(green) closed squares represent the data obtained at 122 62and 37 120583T respectively Apparently the119879
1relaxation time has
a weak dependency on 120588119862 and the 119879
1contrast 120575 exhibits a
weak variance as the magnetic field increases The estimated120575 is 33 times 10minus4 at 37 120583T 63 times 10minus4 at 62120583T and minus94 times 10minus4 at122 120583T which do not appear to bemeaningful compared withstandard deviations of the data Possible reasons for this willbe discussed below
From the mucinous carcinoma tissue (Specimen numberS13-09) a clearly distinguishable NMR signal was measuredas shown in Figure 6 The NMR spectrum has a full-widthhalf-maximum (FWHM) of 17Hz which is about onequarter of the FWHMs of the other tissue which are typically7Hz As a result 120575 is higher than 4 As in the case of high-field MRI [31] mucinous carcinoma tissue can be identifiedunambiguously in low field MRI via 119879
1or 1198792weighted
imaging methods
4 Discussion
In the multiphases relaxation model [27] (5) is based onthe assumption that water molecules in cancer and normalcells can be exchanged rapidly compared with the relaxation
BioMed Research International 7
20 30 40 50 60 70 80 90
Ratio of cancer ()
2
1
0
T1
NT
1T
Figure 5 The 1198791ratio between the tumor (T) and the normal (N)
tissue as a function of the cancer ratio Since the normal tissue inspecimen S13-05 was 5 cancerous the surface ratio of the cancerof this specimen was denoted as 55 The (red) closed circles(blue) open circles and (green) closed squares represent the dataobtained at 122 62 and 37 120583T respectively The error bars indicatethe standard errors The lines are linear fits to data obtained at eachfield
190 200 210 220 230000
005
010
015
020
025
Nor
mal
ized
FFT
(au
)
Frequency (Hz)
FWHM 7Hz
FWHM 17Hz
Figure 6 The normalized spectra for the tumor tissues of S13-09(red) and S13-10 (green) obtained by the experimental conditions of119905delay = 10ms and 2nd 119861
119901= 62 120583T The spectra were normalized
with each spectrum area
rates of water molecules belonging to separate phases in thesystem If this assumption is not valid the 119879
1relaxation of
tissue should exhibitmultiple exponentialsThe1198791values dis-
played in Table 1 are mostly lt100msTherefore the exchangerate between water molecules in cancer and normal cellsshould be at least 10 sminus1 in order for (5) to be applied Suchan exchange rate might not be a realistic value We speculatethat cancer cells are still isolated from normal cells for longerperiods than the time scale of 100ms In this case multipleexponential fitting of the decay curve should be required tocorrectly analyze tissue samples having a cancer ratio in therange of 20sim80The relaxometry of a single (or finite) voxel
in the MR imaging of tumor tissue will therefore be a betterapproach to distinguishing cancer volumes from normal cellswhen they are partially mixed as in our case
Experimental attempts to obtain the optimal 1198791contrast
between cancerous and healthy tissue can be supportedby microscopic physiology studies on cancer cells As theWarburg effect states [32] cancer cells producemostly lactateat the end of their metabolic cycles while normal cellsmostly produce ATPs as an energy source The differentiatedby-products of the metabolic cycle may involve differentinfluences on the relaxation mechanism of water moleculesnear to those chemicals For instance the relaxation rate ofwatermolecules in protein is affected by themolecular weightof the protein [33] Justification for applying a low magneticfield rather than a high field will finally come only after abetter understanding of the relaxation process between by-products of metabolites and water molecules in cells To thisend further physiology-orientated studies at low magneticfields should be conducted in the future
The limitations in the interpretation of our results existin the present work First the measured 119879
1values of several
tissues in Table 1 exhibit rather large variation in the esti-mated errors Currently we suspect that the long averagingtime typically 20min for a single 119879
1measurement may
cause thermally induced problems such as slight variance ordegradation of 119861
119901due to the heated coil That is reasonable if
the circulation of LN2 is not smooth in an instant due to thelocal heating of 119861
119901coil although there is no heat dissipation
at the sample position by using a specially designed 119861119901
coil system Therefore it is definitely required to reduce theaveraging time in order to minimize the unwanted thermaleffects during the measurements Second the comparisonbetween benign and malignant tissues is missing In thiswork malignant tissues are compared only with normaltissues Third the fifteen pairs of tissues investigated inthis work may not seem sufficient for a reliable statisticalanalysis Particularly the result obtained from the mucinoustissue should be confirmed by investigating more tissuesalthough mucinous type is rare Finally the collected tissueswere frozen and preserved in LN2 dewar before the 119879
1
measurementsThe freezing and thawing processmay cause adamage in the tissues In future study amethod of preformingrelaxation measurement before freezing is needed to bedevised
In terms of SNR we believe two points can be improvedwhich will render single voxel relaxometry of breast cancerfeasible First the environmental noise of our system includ-ing SQUID was measured to be 23 fTHz12 and this valueis relatively higher than the other grouprsquos noise environmentbelow 1 fTHz12 [13 25] Second the SNR will increase ifthe 119861119901strength is increased by optimizing the coil system
Currently it was limited to a maximum of 60mT due to heatdissipation from the coil
5 Conclusions
In this study we performed 1198791relaxation measurements
of ex-vivo breast cancer tissue samples in order to verifythe feasibility of confirmative MR imaging of breast cancer
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
BioMed Research International 5
Table1Analyzedandmeasuredparameters
Specim
ennu
mber
Type
Cancer
1198791at122120583
T(m
s)1198791at62120583T(m
s)1198791at37120583T(m
s)Tu
mor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
Tumor
tissue
Normaltissue
S13-01
Ductal
900
604plusmn457
482plusmn264
841plusmn219
914plusmn422
1095plusmn429
69plusmn3
S13-02
aDuctal
250
1611plusmn532
813plusmn247
535plusmn108
941plusmn332
79plusmn352
564plusmn193
S13-03
Ductal
750
687plusmn99
74plusmn385
634plusmn29
623plusmn205
668plusmn137
977plusmn184
S13-04
aLo
bular
600
81plusmn201
546plusmn11
618plusmn5
608plusmn5
595plusmn89
66plusmn84
S13-05
aDuctal+
lobu
lar
605
1001plusmn
74mdash
121plusmn
533
866plusmn5
685plusmn104
105plusmn124
S13-06
Ductal
700
722plusmn228
673plusmn75
636plusmn138
676plusmn83
794plusmn52
634plusmn53
S13-07
Ductal
750
792plusmn58
1013plusmn112
736plusmn48
744plusmn123
682plusmn48
742plusmn64
S13-08
Ductal
900
1692plusmn147
833plusmn156
819plusmn283
829plusmn93
641plusmn184
612plusmn128
S13-09
Mucinou
s100
03477plusmn249
775plusmn269
3586plusmn258
847plusmn134
mdash46
3plusmn56
S13-10
Ductal
750
67plusmn3
635plusmn97
689plusmn27
844plusmn19
656plusmn24
659plusmn56
S13-11
aDuctal
900
861plusmn79
984plusmn298
84plusmn61
603plusmn44
3765plusmn13
876plusmn278
S13-12
Ductal
700
115plusmn117
1006plusmn85
912plusmn33
812plusmn27
498plusmn71
255plusmn78
S13-13
Ductal
900
655plusmn03
758plusmn212
522plusmn154
773plusmn81
64plusmn31
816plusmn145
a Thep
reop
erativec
hemotherapy
was
perfo
rmed
6 BioMed Research International
(a) (b)
(c)
Figure 4 Photomicrographs of (a) the invasive ductal carcinoma sample (Specimen number S13-10) (b) the mucinous carcinoma sample(Specimen number S13-09) and (c) normal breast tissue (Specimen number S13-09) The arrows on the right- and left-hand sides of thepicture in (a) represent irregularly infiltrating epithelia cell clusters and a well-organized normal breast lobule respectively Each specimenwas stained with hematoxylin-eosin and observed at times40 magnification
receiving preoperative chemotherapy one case (Specimennumber S13-02) responded to treatment and the majority ofthe tumor tissue was replaced with fibrotic stroma
Since the tumor tissues contain partial cancers the 1198791
values estimated from the fitting do not yield meaningfulinformation about the 119879
1contrast unless properly analyzed
For correct estimation of the 1198791contrast between cancer and
normal tissue the cancer volume ratio must be considered aspart of the analysis The model devised by Zimmerman andBrittin [27] describes the relaxation phenomenon in a systemof multiple phases According to this model the exchangerate between the twomedia with different119879
1relaxation times
strongly influences the 1198791relaxation process We presume
that the exchange rate of the water molecules betweennormal and cancer cells in a tumor tissue is reasonablyrapid compared with the relaxation rate 1119879
1 Then a single
exponential decay with the average relaxation rate 11198791119879=
1205881198621198791119862+ 1205881198731198791119873 will be obtained in which 119879
1119862and 119879
1119873
are the 1198791relaxation times of the cancer and normal cells
respectively The 120588119862and 120588
119873are volume ratios of the cancer
and normal cells in the tissueThe ratio between1198791119873
and1198791119879
that is 11987911198731198791119879 of each tumor tissue should show a linear
dependence on the 120588119862and the coefficient is the 119879
1contrast
120575 between the cancer and normal cells as
1198791119873
1198791119879
= 1 + 120588119862(1198791119873
1198791119862
minus 1) = 1 + 120588119862120575 (5)
Figure 5 shows the 1198791ratios between the tumor and the
healthy tissue (11987911198731198791119879) as a function of the relative cancer
volume 120588119862 The (red) closed circles (blue) open circles and
(green) closed squares represent the data obtained at 122 62and 37 120583T respectively Apparently the119879
1relaxation time has
a weak dependency on 120588119862 and the 119879
1contrast 120575 exhibits a
weak variance as the magnetic field increases The estimated120575 is 33 times 10minus4 at 37 120583T 63 times 10minus4 at 62120583T and minus94 times 10minus4 at122 120583T which do not appear to bemeaningful compared withstandard deviations of the data Possible reasons for this willbe discussed below
From the mucinous carcinoma tissue (Specimen numberS13-09) a clearly distinguishable NMR signal was measuredas shown in Figure 6 The NMR spectrum has a full-widthhalf-maximum (FWHM) of 17Hz which is about onequarter of the FWHMs of the other tissue which are typically7Hz As a result 120575 is higher than 4 As in the case of high-field MRI [31] mucinous carcinoma tissue can be identifiedunambiguously in low field MRI via 119879
1or 1198792weighted
imaging methods
4 Discussion
In the multiphases relaxation model [27] (5) is based onthe assumption that water molecules in cancer and normalcells can be exchanged rapidly compared with the relaxation
BioMed Research International 7
20 30 40 50 60 70 80 90
Ratio of cancer ()
2
1
0
T1
NT
1T
Figure 5 The 1198791ratio between the tumor (T) and the normal (N)
tissue as a function of the cancer ratio Since the normal tissue inspecimen S13-05 was 5 cancerous the surface ratio of the cancerof this specimen was denoted as 55 The (red) closed circles(blue) open circles and (green) closed squares represent the dataobtained at 122 62 and 37 120583T respectively The error bars indicatethe standard errors The lines are linear fits to data obtained at eachfield
190 200 210 220 230000
005
010
015
020
025
Nor
mal
ized
FFT
(au
)
Frequency (Hz)
FWHM 7Hz
FWHM 17Hz
Figure 6 The normalized spectra for the tumor tissues of S13-09(red) and S13-10 (green) obtained by the experimental conditions of119905delay = 10ms and 2nd 119861
119901= 62 120583T The spectra were normalized
with each spectrum area
rates of water molecules belonging to separate phases in thesystem If this assumption is not valid the 119879
1relaxation of
tissue should exhibitmultiple exponentialsThe1198791values dis-
played in Table 1 are mostly lt100msTherefore the exchangerate between water molecules in cancer and normal cellsshould be at least 10 sminus1 in order for (5) to be applied Suchan exchange rate might not be a realistic value We speculatethat cancer cells are still isolated from normal cells for longerperiods than the time scale of 100ms In this case multipleexponential fitting of the decay curve should be required tocorrectly analyze tissue samples having a cancer ratio in therange of 20sim80The relaxometry of a single (or finite) voxel
in the MR imaging of tumor tissue will therefore be a betterapproach to distinguishing cancer volumes from normal cellswhen they are partially mixed as in our case
Experimental attempts to obtain the optimal 1198791contrast
between cancerous and healthy tissue can be supportedby microscopic physiology studies on cancer cells As theWarburg effect states [32] cancer cells producemostly lactateat the end of their metabolic cycles while normal cellsmostly produce ATPs as an energy source The differentiatedby-products of the metabolic cycle may involve differentinfluences on the relaxation mechanism of water moleculesnear to those chemicals For instance the relaxation rate ofwatermolecules in protein is affected by themolecular weightof the protein [33] Justification for applying a low magneticfield rather than a high field will finally come only after abetter understanding of the relaxation process between by-products of metabolites and water molecules in cells To thisend further physiology-orientated studies at low magneticfields should be conducted in the future
The limitations in the interpretation of our results existin the present work First the measured 119879
1values of several
tissues in Table 1 exhibit rather large variation in the esti-mated errors Currently we suspect that the long averagingtime typically 20min for a single 119879
1measurement may
cause thermally induced problems such as slight variance ordegradation of 119861
119901due to the heated coil That is reasonable if
the circulation of LN2 is not smooth in an instant due to thelocal heating of 119861
119901coil although there is no heat dissipation
at the sample position by using a specially designed 119861119901
coil system Therefore it is definitely required to reduce theaveraging time in order to minimize the unwanted thermaleffects during the measurements Second the comparisonbetween benign and malignant tissues is missing In thiswork malignant tissues are compared only with normaltissues Third the fifteen pairs of tissues investigated inthis work may not seem sufficient for a reliable statisticalanalysis Particularly the result obtained from the mucinoustissue should be confirmed by investigating more tissuesalthough mucinous type is rare Finally the collected tissueswere frozen and preserved in LN2 dewar before the 119879
1
measurementsThe freezing and thawing processmay cause adamage in the tissues In future study amethod of preformingrelaxation measurement before freezing is needed to bedevised
In terms of SNR we believe two points can be improvedwhich will render single voxel relaxometry of breast cancerfeasible First the environmental noise of our system includ-ing SQUID was measured to be 23 fTHz12 and this valueis relatively higher than the other grouprsquos noise environmentbelow 1 fTHz12 [13 25] Second the SNR will increase ifthe 119861119901strength is increased by optimizing the coil system
Currently it was limited to a maximum of 60mT due to heatdissipation from the coil
5 Conclusions
In this study we performed 1198791relaxation measurements
of ex-vivo breast cancer tissue samples in order to verifythe feasibility of confirmative MR imaging of breast cancer
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
6 BioMed Research International
(a) (b)
(c)
Figure 4 Photomicrographs of (a) the invasive ductal carcinoma sample (Specimen number S13-10) (b) the mucinous carcinoma sample(Specimen number S13-09) and (c) normal breast tissue (Specimen number S13-09) The arrows on the right- and left-hand sides of thepicture in (a) represent irregularly infiltrating epithelia cell clusters and a well-organized normal breast lobule respectively Each specimenwas stained with hematoxylin-eosin and observed at times40 magnification
receiving preoperative chemotherapy one case (Specimennumber S13-02) responded to treatment and the majority ofthe tumor tissue was replaced with fibrotic stroma
Since the tumor tissues contain partial cancers the 1198791
values estimated from the fitting do not yield meaningfulinformation about the 119879
1contrast unless properly analyzed
For correct estimation of the 1198791contrast between cancer and
normal tissue the cancer volume ratio must be considered aspart of the analysis The model devised by Zimmerman andBrittin [27] describes the relaxation phenomenon in a systemof multiple phases According to this model the exchangerate between the twomedia with different119879
1relaxation times
strongly influences the 1198791relaxation process We presume
that the exchange rate of the water molecules betweennormal and cancer cells in a tumor tissue is reasonablyrapid compared with the relaxation rate 1119879
1 Then a single
exponential decay with the average relaxation rate 11198791119879=
1205881198621198791119862+ 1205881198731198791119873 will be obtained in which 119879
1119862and 119879
1119873
are the 1198791relaxation times of the cancer and normal cells
respectively The 120588119862and 120588
119873are volume ratios of the cancer
and normal cells in the tissueThe ratio between1198791119873
and1198791119879
that is 11987911198731198791119879 of each tumor tissue should show a linear
dependence on the 120588119862and the coefficient is the 119879
1contrast
120575 between the cancer and normal cells as
1198791119873
1198791119879
= 1 + 120588119862(1198791119873
1198791119862
minus 1) = 1 + 120588119862120575 (5)
Figure 5 shows the 1198791ratios between the tumor and the
healthy tissue (11987911198731198791119879) as a function of the relative cancer
volume 120588119862 The (red) closed circles (blue) open circles and
(green) closed squares represent the data obtained at 122 62and 37 120583T respectively Apparently the119879
1relaxation time has
a weak dependency on 120588119862 and the 119879
1contrast 120575 exhibits a
weak variance as the magnetic field increases The estimated120575 is 33 times 10minus4 at 37 120583T 63 times 10minus4 at 62120583T and minus94 times 10minus4 at122 120583T which do not appear to bemeaningful compared withstandard deviations of the data Possible reasons for this willbe discussed below
From the mucinous carcinoma tissue (Specimen numberS13-09) a clearly distinguishable NMR signal was measuredas shown in Figure 6 The NMR spectrum has a full-widthhalf-maximum (FWHM) of 17Hz which is about onequarter of the FWHMs of the other tissue which are typically7Hz As a result 120575 is higher than 4 As in the case of high-field MRI [31] mucinous carcinoma tissue can be identifiedunambiguously in low field MRI via 119879
1or 1198792weighted
imaging methods
4 Discussion
In the multiphases relaxation model [27] (5) is based onthe assumption that water molecules in cancer and normalcells can be exchanged rapidly compared with the relaxation
BioMed Research International 7
20 30 40 50 60 70 80 90
Ratio of cancer ()
2
1
0
T1
NT
1T
Figure 5 The 1198791ratio between the tumor (T) and the normal (N)
tissue as a function of the cancer ratio Since the normal tissue inspecimen S13-05 was 5 cancerous the surface ratio of the cancerof this specimen was denoted as 55 The (red) closed circles(blue) open circles and (green) closed squares represent the dataobtained at 122 62 and 37 120583T respectively The error bars indicatethe standard errors The lines are linear fits to data obtained at eachfield
190 200 210 220 230000
005
010
015
020
025
Nor
mal
ized
FFT
(au
)
Frequency (Hz)
FWHM 7Hz
FWHM 17Hz
Figure 6 The normalized spectra for the tumor tissues of S13-09(red) and S13-10 (green) obtained by the experimental conditions of119905delay = 10ms and 2nd 119861
119901= 62 120583T The spectra were normalized
with each spectrum area
rates of water molecules belonging to separate phases in thesystem If this assumption is not valid the 119879
1relaxation of
tissue should exhibitmultiple exponentialsThe1198791values dis-
played in Table 1 are mostly lt100msTherefore the exchangerate between water molecules in cancer and normal cellsshould be at least 10 sminus1 in order for (5) to be applied Suchan exchange rate might not be a realistic value We speculatethat cancer cells are still isolated from normal cells for longerperiods than the time scale of 100ms In this case multipleexponential fitting of the decay curve should be required tocorrectly analyze tissue samples having a cancer ratio in therange of 20sim80The relaxometry of a single (or finite) voxel
in the MR imaging of tumor tissue will therefore be a betterapproach to distinguishing cancer volumes from normal cellswhen they are partially mixed as in our case
Experimental attempts to obtain the optimal 1198791contrast
between cancerous and healthy tissue can be supportedby microscopic physiology studies on cancer cells As theWarburg effect states [32] cancer cells producemostly lactateat the end of their metabolic cycles while normal cellsmostly produce ATPs as an energy source The differentiatedby-products of the metabolic cycle may involve differentinfluences on the relaxation mechanism of water moleculesnear to those chemicals For instance the relaxation rate ofwatermolecules in protein is affected by themolecular weightof the protein [33] Justification for applying a low magneticfield rather than a high field will finally come only after abetter understanding of the relaxation process between by-products of metabolites and water molecules in cells To thisend further physiology-orientated studies at low magneticfields should be conducted in the future
The limitations in the interpretation of our results existin the present work First the measured 119879
1values of several
tissues in Table 1 exhibit rather large variation in the esti-mated errors Currently we suspect that the long averagingtime typically 20min for a single 119879
1measurement may
cause thermally induced problems such as slight variance ordegradation of 119861
119901due to the heated coil That is reasonable if
the circulation of LN2 is not smooth in an instant due to thelocal heating of 119861
119901coil although there is no heat dissipation
at the sample position by using a specially designed 119861119901
coil system Therefore it is definitely required to reduce theaveraging time in order to minimize the unwanted thermaleffects during the measurements Second the comparisonbetween benign and malignant tissues is missing In thiswork malignant tissues are compared only with normaltissues Third the fifteen pairs of tissues investigated inthis work may not seem sufficient for a reliable statisticalanalysis Particularly the result obtained from the mucinoustissue should be confirmed by investigating more tissuesalthough mucinous type is rare Finally the collected tissueswere frozen and preserved in LN2 dewar before the 119879
1
measurementsThe freezing and thawing processmay cause adamage in the tissues In future study amethod of preformingrelaxation measurement before freezing is needed to bedevised
In terms of SNR we believe two points can be improvedwhich will render single voxel relaxometry of breast cancerfeasible First the environmental noise of our system includ-ing SQUID was measured to be 23 fTHz12 and this valueis relatively higher than the other grouprsquos noise environmentbelow 1 fTHz12 [13 25] Second the SNR will increase ifthe 119861119901strength is increased by optimizing the coil system
Currently it was limited to a maximum of 60mT due to heatdissipation from the coil
5 Conclusions
In this study we performed 1198791relaxation measurements
of ex-vivo breast cancer tissue samples in order to verifythe feasibility of confirmative MR imaging of breast cancer
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
BioMed Research International 7
20 30 40 50 60 70 80 90
Ratio of cancer ()
2
1
0
T1
NT
1T
Figure 5 The 1198791ratio between the tumor (T) and the normal (N)
tissue as a function of the cancer ratio Since the normal tissue inspecimen S13-05 was 5 cancerous the surface ratio of the cancerof this specimen was denoted as 55 The (red) closed circles(blue) open circles and (green) closed squares represent the dataobtained at 122 62 and 37 120583T respectively The error bars indicatethe standard errors The lines are linear fits to data obtained at eachfield
190 200 210 220 230000
005
010
015
020
025
Nor
mal
ized
FFT
(au
)
Frequency (Hz)
FWHM 7Hz
FWHM 17Hz
Figure 6 The normalized spectra for the tumor tissues of S13-09(red) and S13-10 (green) obtained by the experimental conditions of119905delay = 10ms and 2nd 119861
119901= 62 120583T The spectra were normalized
with each spectrum area
rates of water molecules belonging to separate phases in thesystem If this assumption is not valid the 119879
1relaxation of
tissue should exhibitmultiple exponentialsThe1198791values dis-
played in Table 1 are mostly lt100msTherefore the exchangerate between water molecules in cancer and normal cellsshould be at least 10 sminus1 in order for (5) to be applied Suchan exchange rate might not be a realistic value We speculatethat cancer cells are still isolated from normal cells for longerperiods than the time scale of 100ms In this case multipleexponential fitting of the decay curve should be required tocorrectly analyze tissue samples having a cancer ratio in therange of 20sim80The relaxometry of a single (or finite) voxel
in the MR imaging of tumor tissue will therefore be a betterapproach to distinguishing cancer volumes from normal cellswhen they are partially mixed as in our case
Experimental attempts to obtain the optimal 1198791contrast
between cancerous and healthy tissue can be supportedby microscopic physiology studies on cancer cells As theWarburg effect states [32] cancer cells producemostly lactateat the end of their metabolic cycles while normal cellsmostly produce ATPs as an energy source The differentiatedby-products of the metabolic cycle may involve differentinfluences on the relaxation mechanism of water moleculesnear to those chemicals For instance the relaxation rate ofwatermolecules in protein is affected by themolecular weightof the protein [33] Justification for applying a low magneticfield rather than a high field will finally come only after abetter understanding of the relaxation process between by-products of metabolites and water molecules in cells To thisend further physiology-orientated studies at low magneticfields should be conducted in the future
The limitations in the interpretation of our results existin the present work First the measured 119879
1values of several
tissues in Table 1 exhibit rather large variation in the esti-mated errors Currently we suspect that the long averagingtime typically 20min for a single 119879
1measurement may
cause thermally induced problems such as slight variance ordegradation of 119861
119901due to the heated coil That is reasonable if
the circulation of LN2 is not smooth in an instant due to thelocal heating of 119861
119901coil although there is no heat dissipation
at the sample position by using a specially designed 119861119901
coil system Therefore it is definitely required to reduce theaveraging time in order to minimize the unwanted thermaleffects during the measurements Second the comparisonbetween benign and malignant tissues is missing In thiswork malignant tissues are compared only with normaltissues Third the fifteen pairs of tissues investigated inthis work may not seem sufficient for a reliable statisticalanalysis Particularly the result obtained from the mucinoustissue should be confirmed by investigating more tissuesalthough mucinous type is rare Finally the collected tissueswere frozen and preserved in LN2 dewar before the 119879
1
measurementsThe freezing and thawing processmay cause adamage in the tissues In future study amethod of preformingrelaxation measurement before freezing is needed to bedevised
In terms of SNR we believe two points can be improvedwhich will render single voxel relaxometry of breast cancerfeasible First the environmental noise of our system includ-ing SQUID was measured to be 23 fTHz12 and this valueis relatively higher than the other grouprsquos noise environmentbelow 1 fTHz12 [13 25] Second the SNR will increase ifthe 119861119901strength is increased by optimizing the coil system
Currently it was limited to a maximum of 60mT due to heatdissipation from the coil
5 Conclusions
In this study we performed 1198791relaxation measurements
of ex-vivo breast cancer tissue samples in order to verifythe feasibility of confirmative MR imaging of breast cancer
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
8 BioMed Research International
Since the volumes of the investigated tissues were smallimprovements on our ULF-NMR setup and signal processingapproach were required A specially designed 119861
119901coil system
including liquid nitrogen cooling and thermal isolationproduces a reliable high field without heat dissipation atthe sample position In addition noise elimination usingBayesian statistics enabled us to overcome a relatively smallSNR In spite of those efforts the 119879
1relaxometry measured
at 37 62 and 122 120583T failed to exhibit a meaningful fingerprintexpected at low magnetic fields At this moment howeverwe can not conclude that the absence of the fingerprint is anintrinsic characteristic of ductal- and lobular-type breast can-cer tissues owing to the remaining limitations discussed inthe previous section Therefore further improvements of thesystems andmethods are required to obtain valuable outputsIn contrast mucinous cancer showed a reliable enhancementin 1198791contrast which would be a useful application of ULF-
MRIThis work is the first experimental approach toward thediscrimination of breast cancer tissues by 119879
1contrasts at low
fields The equipment and methodologies that we developedin this study are believed to be a useful guide to future work
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Authorsrsquo Contribution
Seong-Joo Lee and Jeong Hyun Shim contributed equally tothis work
Acknowledgment
This work was supported by the World Class Laboratory(WCL)Grant fromKorea Research Institute of Standards andScience
References
[1] K Kerlikowske D Grady J Barclay E A Sickles and VErnster ldquoEffect of age breast density and family history on thesensitivity of first screening mammographyrdquoThe Journal of theAmerican Medical Association vol 276 no 1 pp 33ndash38 1996
[2] U Fischer L Kopka and E Grabbe ldquoBreast carcinoma effect ofpreoperative contrast-enhancedMR imaging on the therapeuticapproachrdquo Radiology vol 213 no 3 pp 881ndash888 1999
[3] H C ShinWHan H GMoon et al ldquoLimited value and utilityof breast MRI in patients undergoing breast-conserving cancersurgeryrdquo Annals of Surgical Oncology vol 19 no 8 pp 2572ndash2579 2012
[4] ldquoBreast MRI for screeningrdquo httpwwwbreastcancerorgsymptomstestingtypesmriscreening
[5] J-P Korb and R G Bryant ldquoMagnetic field dependence ofproton spin-lattice relaxation timesrdquo Magnetic Resonance inMedicine vol 48 no 1 pp 21ndash26 2002
[6] R Kimmich and E Anoardo ldquoField-cyclingNMR relaxometryrdquoProgress in Nuclear Magnetic Resonance Spectroscopy vol 44no 3-4 pp 257ndash320 2004
[7] B Halle ldquoMolecular theory of field-dependent proton spin-lattice relaxation in tissuerdquoMagnetic Resonance inMedicine vol56 no 1 pp 60ndash72 2006
[8] K-W Huang H-H Chen H-C Yang et al ldquoDiscriminat-ing hepatocellular carcinoma in rats using a high-T
119888SQUID
detected nuclear resonance spectrometer in a magnetic shield-ing boxrdquo PLoS ONE vol 7 no 10 Article ID e47057 2012
[9] S H Koenig and R D Brown III ldquoField-cycling relaxometryof protein solutions and tissue implications for MRIrdquo Progressin Nuclear Magnetic Resonance Spectroscopy vol 22 no 6 pp487ndash567 1990
[10] S K Lee M Moszligle W Myers et al ldquoSQUID-detected MRI at132 120583T with T
1-weighted contrast established at 10120583T-300mTrdquo
Magnetic Resonance in Medicine vol 53 no 1 pp 9ndash14 2005[11] J Clarke M Hatridge and M Moszligle ldquoSQUID-detected mag-
netic resonance imaging in microtesla fieldsrdquo Annual Review ofBiomedical Engineering vol 9 no 1 pp 389ndash413 2007
[12] S-H Liao K-W Huang H-C Yang et al ldquoCharacterization oftumors using high-T
119888superconducting quantum interference
device-detected nuclear magnetic resonance and imagingrdquoApplied Physics Letters vol 97 no 26 Article ID 263701 2010
[13] S Busch M Hatridge M Moszligle et al ldquoMeasurements ofT1-relaxation in ex vivo prostate tissue at 132 120583Trdquo Magnetic
Resonance in Medicine vol 67 no 4 pp 1138ndash1145 2012[14] S-J Lee K Kim C S Kang S-M Hwang and Y-H Lee ldquoPre-
polarization enhancement by dynamic nuclear polarizationin SQUID-based ultra-low-field nuclear magnetic resonancerdquoSuperconductor Science and Technology vol 23 no 11 ArticleID 115008 2010
[15] C S Kang K Kim S-J Lee et al ldquoApplication of the doublerelaxation oscillation superconducting quantum interferencedevice sensor to micro-tesla 1H nuclear magnetic resonanceexperimentsrdquo Journal of Applied Physics vol 110 no 5 ArticleID 053906 2011
[16] JH Shim S-J Lee K-K Yu S-MHwang andKKim ldquoStrongpulsed excitations using circularly polarized fields for ultra-lowfield NMRrdquo Journal of Magnetic Resonance vol 239 no 1 pp87ndash90 2014
[17] S-M Hwang K Kim K K Yu et al ldquoType-I superconductorpick-up coil in superconducting quantum interference device-based ultra-low field nuclear magnetic resonancerdquo AppliedPhysics Letters vol 104 no 6 pp 062602ndash062602-4 2014
[18] K Kim S-J Lee C S Kang S-M Hwang Y-H Lee and K KYu ldquoToward a brain functional connectivity mapping modalityby simultaneous imaging of coherent brainwavesrdquoNeuroImagevol 91 no 1 pp 63ndash69 2014
[19] J H Shim S-J Lee S-M Hwang K K Yu and K Kim ldquoTwo-dimensional NMR spectroscopy of 13C methanol at less than5 120583Trdquo Journal of Magnetic Resonance vol 246 no 1 pp 4ndash82014
[20] R McDermott A H Trabesinger M Muck E L Hahn APines and J Clarke ldquoLiquid-state NMR and scalar couplings inmicroteslamagnetic fieldsrdquo Science vol 295 no 5563 pp 2247ndash2249 2002
[21] V S Zotev ANMatlashov P L Volegov et al ldquoMicroteslaMRIof the human brain combined with MEGrdquo Journal of MagneticResonance vol 194 no 1 pp 115ndash120 2008
[22] P T Vesanen J O Nieminen K C J Zevenhoven et alldquoHybrid ultra-low-field MRI and magnetoencephalographysystem based on a commercial whole-head neuromagnetome-terrdquo Magnetic Resonance in Medicine vol 69 no 6 pp 1795ndash1804 2013
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
BioMed Research International 9
[23] R H Kraus Jr P Volegov A Matlachov and M Espy ldquoTowarddirect neural current imaging by resonant mechanisms at ultra-low fieldrdquo NeuroImage vol 39 no 1 pp 310ndash317 2008
[24] M Burghoff H-H Albrecht S Hartwig et al ldquoOn the feasi-bility of neurocurrent imaging by low-field nuclear magneticresonancerdquo Applied Physics Letters vol 96 no 23 Article ID233701 2010
[25] R Korber J O Nieminen N Hofner et al ldquoAn advanced phan-tom study assessing the feasibility of neuronal current imagingby ultra-low-field NMRrdquo Journal of Magnetic Resonance vol237 no 1 pp 182ndash190 2013
[26] K Kim ldquoToward cardiac electrophysiological mapping basedonmicro-TeslaNMR a novelmodality for localizing the cardiacreentryrdquo AIP Advances vol 2 no 2 Article ID 022156 2012
[27] J R Zimmerman and W E Brittin ldquoNuclear magnetic res-onance studies in multiple phase systems lifetime of a watermolecule in an adsorbing phase on silica gelrdquo The Journal ofPhysical Chemistry vol 61 no 10 pp 1328ndash1333 1957
[28] W Franzen ldquoGeneration of uniform magnetic fields by meansof air-core coilsrdquo Review of Scientific Instruments vol 33 no 9pp 933ndash938 1962
[29] N IMatter G C Scott T Grafendorfer AMacovski and SMConolly ldquoRapid polarizing field cycling in magnetic resonanceimagingrdquo IEEE Transactions on Medical Imaging vol 25 no 1pp 84ndash93 2006
[30] J C Hoch and A S Stern NMR Data Processing Wiley-LissNew York NY USA 1996
[31] M Kawashima Y Tamaki T Nonaka et al ldquoMR imagingof mucinous carcinoma of the breastrdquo American Journal ofRoentgenology vol 179 no 1 pp 179ndash183 2002
[32] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009
[33] J Zhong J C Gore and IM Armitage ldquoQuantitative studies ofhydrodynamic effects and cross-relaxation in protein solutionsand tissues with proton and deuteron longitudinal relaxationtimesrdquo Magnetic Resonance in Medicine vol 13 no 2 pp 192ndash203 1990
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Anatomy Research International
PeptidesInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
International Journal of
Volume 2014
Zoology
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Molecular Biology International
GenomicsInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioinformaticsAdvances in
Marine BiologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Signal TransductionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
Evolutionary BiologyInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biochemistry Research International
ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Genetics Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Advances in
Virolog y
Hindawi Publishing Corporationhttpwwwhindawicom
Nucleic AcidsJournal of
Volume 2014
Stem CellsInternational
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Enzyme Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Microbiology
Top Related