www.appliedradiology.com APPLIED RADIOLOGY©� 9May 2006

The imaging of brain tumors hassignificantly improved with theuse of advanced magnetic reso-

nance (MR) techniques, such as spec-troscopy, perfusion, and functionalimaging. Conventional MR imaging(MRI) provides mainly anatomic orstructural information about the brain.Unlike conventional imaging, advancedMR techniques also provide physiologi-cal information concerning metabolismand hemodynamics. These techniquesnot only aid in the imaging diagnosis ofbrain tumors, they may also play a rolein clinical management of patients withbrain tumors. A concise compendium ofthe physiology, techniques, and clinicalapplications of MR perfusion imaging,spectroscopy, functional MR imaging(fMRI), and diffusion tensor imaging(DTI) in the setting of neuro-oncology isreviewed in this article.

Perfusion imagingBrain tumors can induce angiogenesis

or the formation of new blood vessels.Hypoxia, which occurs as a tumor out-grows its blood supply, can produce an-giogenic cytokines; these cytokines areresponsible for angiogenesis.1 Tumor

vessels that are produced in this mannerare histologically abnormal and morepermeable than normal. They are alsodisorganized and tortuous.2 These vascu-lar abnormalities and altered flow dynam-ics lead to changes in blood volume andflow, which are exploited in MR perfu-sion imaging. The most frequently usedmeasure of perfusion in neuro-oncologyis the cerebral blood volume (CBV). Thecerebral blood volume (or the volume ofblood passing through a portion of thebrain) is measured in milliliters of bloodper 100 grams of brain tissue (mL/100 g).

The most common perfusion tech-nique is T2* dynamic susceptibilityimaging. The T2* effects of gadoliniumresult in decreased signal intensity duringthe passage of gadolinium. The change in signal intensity is plotted against timeto form a signal intensity time curve (Fig-ure 1). The CBV is estimated from thearea encompassed by the curve, which isinverted in this case, since there is signalloss. Repetitive imaging is performed

shortly before, during, and after the pas-sage of gadolinium. Generally, 0.2 mmol/Kgof gadolinium is injected at a high rateusing a power injector. The CBV is nor-malized to uninvolved portions of thebrain. In cases in which an arterial inputfunction is not determined, only a relativeCBV (rCBV) can be calculated. Dynamicsusceptibility perfusion imaging is basedon the premise that contrast materialremains within the intravascular compart-ment. High permeability or leakiness inregions of marked breakdown of theblood-brain barrier (BBB) results inintravascular gadolinium extravasatinginto the interstitial space. Extravasationcan significantly affect calculations andalter CBV values. Several methods havebeen used to correct—or, more appro-priately, compensate for—the unwantedeffect of extravasation on rCBV calcula-tions, including excluding portions of the signal intensity time curve from thecalculations. This corrective method stillleads to underestimation of the rCBV.

Advanced MR techniques in brain tumor imaging

Sasan Karimi, MD; Nicole M. Petrovich, BA; Kyung K. Peck, PhD; Bob L. Hou, PhD; Andrei I. Holodny, MD

Dr. Karimi is an Assistant Professor,Department of Radiology, Ms. Petrovichis the Senior Neurodiagnostic fMRI Specialist, Department of Radiology,Dr. Peck is an Instructor, and Dr. Houis an Assistant Attending Physician,Department of Medical Physics, and Dr. Holodny is a Professor, Departmentof Radiology, Memorial Sloan-KetteringCancer Center, New York, NY.

10 � APPLIED RADIOLOGY© www.appliedradiology.com May 2006

Another method, sloping baseline, leadsto artifactually high rCBV. Using T1-weighted dynamic perfusion imaging

can eliminate the problem with thebreakdown of the BBB and permeability.In this technique, tumor enhancement or

permeability itself is used to calculate the rCBV, so it does not need to be cor-rected.3

In general, high-grade brain tumorshave greater rCBV than low-grade le-sions (Figure 2).1,2,4,5 MR perfusion canhelp identify and localize higher gradecomponents of tumors in guiding stereo-tactic biopsy and can also provide a non-invasive estimate of tumor grade (Figure3).1-4 Since the enhancing or even the T2borders of gliomas do not represent thetrue margins of the tumor, MR perfusioncan be more sensitive in defining thetrue extent of gliomas than can anatomicMR imaging.6 Better delineation oftumor borders can help in radiation andsurgical planning. In the future, MR per-fusion imaging will probably play a rolein better defining tumor margins forradiation and surgical planning.

Differentiating tumor recurrence fromradiation necrosis is a problem that clini-cians and radiologists face frequently. Itshould be noted that cases of pure tumorrecurrence and radiation necrosis occur inonly a minority of cases. The majority ofcases fall within a spectrum containing amixture of both tumor and necrosis. Thisadds further complexity to reaching a cor-rect diagnosis. Unlike tumors, which haveelevated rCBV, radiation necrosis hasbeen shown to have diminished rCBV(Figure 4).7 Since a significant number ofpatients fall within the spectrum, perfu-sion imaging is of limited value in differ-entiating between the two. The advantageof MR perfusion CBV data is its high positive predictive value for the presenceof high-grade malignancy. In other words,if the CBV maps reveal elevated perfu-sion, there is tumor present. The speci-ficity of the diagnostic task can be in-creased when perfusion imaging is com-bined with MR spectroscopy.

Measurements of rCBV have beenshown to be more useful in assessingresponse to radiotherapy in patientstreated with stereotactic radiosurgery(SRS).8 The vascularity of metastasescan decrease within a few weeks of treat-ment; however, the volume of enhance-ment might not change for severalmonths. MR perfusion imaging seems

BRAIN TUMOR MRI

FIGURE 1. This signal intensity [SI] time curve shows a decrease in SI as gadolinium passesthrough.

FIGURE 2. Biopsy-proven glioblastoma multiforme. (A) An axial T1-weighted postcontrast MRimage shows a heterogeneous enhancing lesion within the posterior right frontal and parietallobes. (B) Increased blood volume in the region of the tumor is shown on the relative cerebralblood volume map.

A B

www.appliedradiology.com APPLIED RADIOLOGY©� 13May 2006

promising in the follow-up of patientswith brain metastasis, but more researchis necessary to evaluate its potential.

SpectroscopyClinical applications of proton mag-

netic resonance spectroscopy (1H-MRS)are increasing as the techniques andhardware have become more robust anduser-friendly. Proton MRS provides bio-chemical and metabolic informationabout tumors and normal brain.9 Theinformation obtained from 1H-MRS isunique and independent of that obtained

from other MRI techniques. In theirstudy of pediatric brain tumors, Tzika etal10 showed that there is no correlationbetween the metabolic profile of tumorsand other imaging parameters, such asenhancement, diffusion, and rCBV.

Spectroscopy can be done in single- ormultivoxel (MRS imaging) forms. The 2most commonly used methods for vol-ume selection/excitation are stimulatedecho acquisition mode (STEAM) andpoint-resolved spectroscopy sequence(PRESS). In general, shorter echo timesare better achieved with STEAM; how-

ever, it is more sensitive to motion.11 Intheory, for the same total echo time, thesignal of PRESS is twice as great as thatof STEAM; PRESS is also less sensitiveto motion. With improving software,PRESS seems to be the most commonlyused method of volume selection in clini-cal practice at present time.

The advantage of single-voxel 1H-MRS is its short acquisition time (approx-imately 5 minutes). The down side is thatit lacks spatial resolution and cannot beused to better define the true extent of aglioma. Histologically, gliomas are het-erogeneous, and, therefore, single-voxelspectroscopy cannot be used to mapregional metabolic variation. In otherwords, single-voxel spectroscopy alonecannot reliably define the highest-gradecomponents of the tumor. It may also suf-fer from significant averaging with adja-cent normal brain tissues.

In the spectroscopy literature, the mostcommonly used echo times are 144 msecand 270 msec. At these long echo times,the spectrum is dominated by 5 differentmetabolite peaks. These are the choline(Cho)-containing compounds, creatine(Cr), N-acetylaspartate (NAA), lactate,and lipid (Figure 5). The choline peakreflects cell membrane turnover. Creatineis a good surrogate for energy synthesis,and NAA is a marker that is exclusive toneuronal cells. Lactate results from anaer-obic metabolism and is detected innecrotic tumors and infarcted tissue. Cel-lular and myelin breakdown productsresult in prominent lipid peaks. In tumors,choline-containing compounds are in-creased, and NAA is decreased relative touninvolved or normal brain tissue.12,13

This pattern of metabolic change is thespectroscopic hallmark of brain tumors.

Combined with MRI, MRS can aid inthe evaluation of tumor type and grade.The higher-grade gliomas tend to exhibithigher Cho/Cr and Cho/NAA ratios. Thehigh-grade gliomas also tend to havelipid and lactate as the result of necrosis(Figure 6).14 MR spectroscopy can helpdifferentiate enhancing tumor from othercauses of enhancement (mainly necrosis)and is more specific in differentiatingnonenhancing tumor from edema and

BRAIN TUMOR MRI

A B

C D

FIGURE 3. (A) The axial T1-weighted postgadolinium image is unremarkable, with no evi-dence of abnormal enhancement. (B) An axial fluid-attenuated inversion recovery image at thesame levels as in (A) shows abnormal hyperintensity in the left peri-insular region, which issuggestive of a low-grade primary tumor, given its lack of enhancement. (C) The relative cere-bral blood volume map shows a subtle asymmetric increase in blood volume in the left subin-sular region (arrow), which suggests a high-grade tumor. (D) The marked elevation of thecholine-to-creatine ratio (arrow) is also suggestive of a high-grade tumor. A biopsy of the lesionconfirmed that it was a high-grade tumor, an anaplastic astrocytoma.

14 � APPLIED RADIOLOGY© www.appliedradiology.com May 2006

other causes of T2 prolongation. Thesequalities have been exploited in order tobetter define the true extent and morphol-ogy of gliomas. This information has thepotential to significantly alter target vol-umes and doses in radiation therapy ofbrain gliomas when compared with con-ventional radiotherapy. Although this isan attractive concept, there are no studiesto show benefits, changes in failure pat-terns, or improved survival. MR spec-troscopy is utilized more and more bydifferent groups in assessing response to

therapy in patients with primary braintumors or metastases.

MR spectroscopy can noninvasivelyenable the distinction between a solitarymetastasis and high-grade gliomas, par-ticularly when combined with perfusion MR imaging. In their study, Law et al15

showed that measurements of Cho andmean rCBV in the perienhancing regionare useful in differentiating solitarymetastases from high-grade gliomas. Inthe perienhancing region, T2 prolonga-tion is partly due to tumor infiltration

(nonenhancing tumor) in patients withhigh-grade gliomas. Whereas in the caseof metastases, the hyperintensity sur-rounding the region of enhancement is dueto vasogenic edema or nonspecific treat-ment effects rather than infiltrating tumor.Therefore, elevated levels of choline and/or rCBV surrounding a peripherally en-hancing mass reflect tumor infiltration in a high-grade glioma (Figure 7).

As with perfusion MRI, MRS is alsouseful in estimating tumor grade. It iscommonly observed that the Cho/Crratio increases with histologic grade.16,17

However, because of the innate hetero-geneity of brain tumors, there is a signif-icant overlap between Cho/Cr levels andtumor grade.

Following treatment, MRS has a lim-ited role in the assessment of patients.Frequently, there is a mixture of tumorand necrosis after therapy. This limits theutility of MRS in differentiating resid-ual/recurrent tumor from radiationnecrosis, as is the case with MR perfu-sion. As a tumor responds to treatment,the choline decreases and lactate and/orlipids may increase.18,19 MR spectroscopycan play a useful role after treatment inassessing the therapeutic response. Thisis particularly important for early detec-tion of treatment failure so that an inef-fective treatment can be modified prior toa significant progression of disease.

Functional MRIFunctional MRI is used for the purpose

of neurosurgical planning and neurologicrisk assessment in the treatment of braintumors.20,21 It localizes the eloquent cor-tices controlling language, motor, andmemory functions. The results of an fMRIstudy can alter a neurosurgical approachto a tumor, suggest that surgery is a safeoption in cases in which it might not oth-erwise have been offered, or steer a clini-cian away from neurosurgery and towardother treatment options when the risk ofdamage to the eloquent cortex is high.

Functional MRI is commonly used tomap language function.23-25 It is used to localize the areas of the cortex work-ing to support speech and to determinehemispheric dominance for language

BRAIN TUMOR MRI

B

FIGURE 4. (A) An axial T1-weighted MR image of a patient with renal cell carcinoma who wastreated with stereotactic radiosurgery for a brain metastasis. The heterogeneous enhancingright insular metastasis was slightly larger following radiotherapy. (B) The relative cerebralblood volume map shows marked decreased blood volume in the region of enhancement,which is consistent with radiation necrosis as the dominant underlying cause of enhancement.The lesion has been stable on subsequent follow-up examinations.

FIGURE 5. (A) A single-voxel spectrum of normal brain shows the choline-to-creatine (Cho/Cr)ratio near unity. (B) A single-voxel spectrum of a glioblastoma multiforme reveals a markedincrease in the Cho/Cr ratio, decreased N-acetylaspartate, and a lactate doublet, which isinverted at long echo time (144 msec).

A

A B

www.appliedradiology.com APPLIED RADIOLOGY©� 15May 2006

(Figure 8). Because fMRI is based on a noninvasive, endogenous signal, it has the potential to replace the invasiveWada test (also known as the intracarotidamytal test), which is the current goldstandard in language and memory lateral-ization.26,27

Functional MRI has revealed unex-pected hemispheric language dominance

in both adults and children.28-30 In a casedescribed by our group,28 a 62-year-oldright-handed man with a mostly nonen-hancing left temporoparietal glioma in theexpected anatomic location of Wenicke’sarea presented without language impair-ment. The fMRI revealed split hemispher-ic dominance for language, with Broca’s area localizing to the left hemisphere and

Wernicke’s to the right hemisphere.Intraoperative electrocortical stimula-tion confirmed the fMRI results, andthe patient underwent gross-total resec-tion. This patient might not otherwisehave been offered an operation had thefMRI not suggested atypical languagelocalization.

Language paradigms vary with thelocation of the tumor. Typically, forfrontal lesions, the patient will performtasks such as generating words to a pre-sented letter or verbs to visually pre-sented nouns, punctuated by periods ofrest. For posterior language localiza-tion, patients can be asked to name pic-tures or fill in the appropriate missingword in a sentence. While targeted test-ing is preferable, often both Broca’s andWernicke’s areas are activated duringboth productive and receptive languagetasks, making it not essential to tailorthe language task to the lesion location.However, it should be noted that poste-rior language function (Wernicke’sarea), like many cognitive tasks, can bedifficult to measure in patients.31 A tar-geted approach may help lessen thevariability in capturing this sometimeselusive language area.

The Wada test produces informationabout both language and memory func-tion. As a result, the ability to use fMRIto noninvasively measure both languageand memory function may further dis-place the Wada test. There is as yet nocommonly accepted set of memory tasksfor the purpose of neurosurgical plan-ning. Golby et al32 use a “novel versusrepeated” protocol through which verbalmemory and spatial memory can bemeasured. In this protocol, the patient is presented with novel and repeatedimages of faces, patterns, scenes, andword pairs (in separate trials). The analy-sis yields those areas that are active dur-ing the novel trials in which the patientencodes the new stimuli. As nonin-vasive, repeatable techniques, such asfMRI, improve in the measurement ofhemispheric dominance for memory andlanguage, invasive tests, with their asso-ciated morbidity (such as the Wada test),will be replaced.

BRAIN TUMOR MRI

FIGURE 6. A biopsy-proven glioblastoma multiforme. (A) An axial T1-weighted MR imageshows an infiltrative enhancing mass within the right frontal lobe with central necrosis. (B) An MRspectroscopic image overlaid on an axial fluid-attenuated inversion recovery image that encom-passes the tumor and surrounding brain tissue is consistent with a necrotic high-grade tumor(magnified representative voxel—lower right-hand corner).

A B

FIGURE 7. (A) An axial postcontrast T1-weighted MR image shows a heterogeneous mass inthe left parietal lobe. (B) A “screen save” from a 3-dimensional MR spectroscopic image isoverlaid onto a magnified fluid-attenuated inversion recovery (FLAIR) scout image, showingan increased choline-to-creatine ratio in the perienhancing region. Note the relative lack ofabnormal FLAIR hyperintensity in the perienhancing region (arrow).

A B

16 � APPLIED RADIOLOGY© www.appliedradiology.com May 2006

BRAIN TUMOR MRI

FIGURE 10. A diffusion tractography image in a normal volunteershows the corticospinal tracts of the hand and the foot bilaterallyextending from the precentral gyrus (the location of the motorhomunculus) to the posterior limb of the internal capsule.

FIGURE 8. A preoperative blood-oxygen-level–dependent func-tional MRI performed with language task shows Broca’s (greenarrow) and Wernecke’s (yellow arrow) areas.

FIGURE 9. A preoperative blood-oxygen-level–dependent functional MRIwith a bilateral finger-tapping paradigm shows the spatial relationshipbetween the tumor and motor activation (green arrow), contralateral motoractivation (green arrow), and supplemental motor activation (yellow arrow).Note the relative decrease in activation on the side of the tumor.

Motor mapping, by contrast, is relatively easy for cognitivelyimpaired patients to perform and produces consistent and reliablefMRI maps.33 Often, patients perform cued movements of the fin-gers, feet, and/or tongue, depending on the location of the lesion(Figure 9).

Localizing the motor strip and coregistering the results to a surgicalscan prior to a neurosurgical intervention can help guide the directcortical stimulation during an awake craniotomy and possibly shortenoperation time. In some cases, using fMRI to confirm the expectedlocation of the motor strip may avoid awake neurosurgery altogether.

Diffusion tensor imaging: A way to see beyond the gray areasAlthough blood-oxygen-level–dependent (BOLD) fMRI is able to

directly visualize the exact location of a functional area of the brainadjacent to a brain tumor,34,35 there is a limitation in that the techniquecan depict activation only in the cortical gray matter. Therefore, theinformation provided to the operating neurosurgeon is limited to thecortex. Brain tumors also invade the white matter,36,37 however, BOLDfMRI is not able to provide accurate information about the location ofmajor white matter tracts. Accidental resection or transection of amajor white matter tract can lead to devastating consequences.

There are 2 main reasons why preoperative identification of whitematter tracts is important. First, accurate localization of importantwhite matter tracts can affect the decision of whether or not to operate.

www.appliedradiology.com APPLIED RADIOLOGY©� 17May 2006

For example, if the main mass of thetumor straddles the corticospinal tract, aneurosurgeon may be averse to even at-tempting a gross total resection. Clearly,knowledge of the location of the corti-cospinal tract in such a case would beadvantageous, since the neurosurgeonwould not have to perform a craniotomy,operate on the brain, and perform directwhite matter stimulation only to discoverthat the tumor is inoperable. Secondly,preoperative localization of importantwhite matter tracts is essential in surgicalplanning. For example, in a tumor thatinvolves the corona radiata, it is oftenvery difficult, if not impossible, to deter-mine the relationship of the tumor to thecorticospinal tract or the thalamocorticaltract. This task becomes even more diffi-cult with the inevitable presence of masseffect and infiltration of normal struc-tures by tumor. Consequently, the neuro-surgeon may assume that the majority ofthe tumor to be resected is anterior to thecorticospinal tract, only to discover dur-ing the operation that in reality, the corti-cospinal tract is posterior to the tumorand that the initial approach taken hasbeen incorrect.

Diffusion tensor imaging38,39 is a newMRI technique that is sensitive to direc-tional movements of water moleculesand that allows the identification of func-tional white matter tracts in vivo. Diffu-sion tensor imaging has the potential toestablish spatial relationships betweeneloquent white matter and tumor bordersand provide clinically valuable informa-tion to assess the progression and regres-sion of white matter tracts as a result oftumor growth or resection.

In gray matter, it is usually sufficientto characterize the diffusion characteris-tics with a single apparent diffusion co-efficient (ADC), because measuredwater diffusivity is largely independentof the orientation of the tissue. However,in an anisotropic area, such as white mat-ter, where the measured diffusivity isknown to depend upon the orientation of the tissue, a single ADC is not able to describe the orientation-dependentwater mobility in the tissue. Diffusiontensor imaging, which is based on the

orientation-dependent water diffusion,can be achieved using a single-shot diffusion-weighted (DW) spin-echo echo-planar imaging (EPI) pulse sequence, inwhich 2 symmetric trapezoidal gradientpulses are added around a 180° refocus-ing pulse in the required gradient chan-nel. Sets of DW-EPI images are collectedwith diffusion gradients applied sequen-tially along at least 6 predetermineddirections. From the DTI data, the 6 inde-pendent elements of the diffusion tensorcan be calculated for each pixel. In tensortheory,38 the directions of the main axesrepresent the so-called eigenvectors andtheir length the so-called eigenvalues ofthe tensor. Diagonalization of the tensorcan be used to calculate the eigenvalues(λ1,λ2,λ3) and the eigenvector. The vec-tor corresponding to the largest eigen-value represents the direction in whichwater diffusion is greatest and is assumedto correspond to the predominant fiberorientation within each voxel. The ele-ments of the tensor are used to yield amean diffusivity map (D) and the frac-tional anisotropy (FA). D is an indicatorof free water fraction. FA is used to mea-sure the fraction of the total magnitude ofD that is anisotropic and has a value of 0for isotropic diffusion (λ1=λ2=λ3) and 1for complete anisotropic diffusion (λ1>0; λ2=λ3=0). Therefore, it describesdeviation from isotropic diffusion.Previous studies37,40 have shown that neo-plasms and surrounding edematous brainhave an increase of free water fraction(high D value) and loss of structuralorganization (reduced FA value).

A number of recent developments haveled to a new application of DWI termeddiffusion tractography. This applicationuses diffusion tensor data to identify spe-cific white matter tracts as opposed towhite matter tracts in general (Figure 10).For example, one can determine the loca-tion of the corticospinal tract or the thala-mocortical tract and differentiate thesetracts from all other white matter tracts inthe corona radiata. The ability to exactlydefine specific tracts traversing the coronaradiata and other nebulous white matterstructures has already had an impact onthe treatment of brain lesions. A number

of recent publications have shown re-markable images of specific white mattertracts in normal subjects using diffusiontractography.41,42 However, this technol-ogy has just made its appearance and willno doubt expand even further.

ConclusionThe integration of advanced imaging

techniques (such as fMRI, spectroscopy,DTI, and perfusion imaging) will offerincreasingly detailed information aboutpathologic processes. Furthermore, it ishoped that such detailed informationwill aid in the development of new treat-ments and will improve our understand-ing of the mechanism underlying neuro-logical disorders.

REFERENCES1. Jackson A, Kassner A, Annesley-Williams D, et al.Abnormalities in the recirculation phase of contrastagent bolus passage in cerebral gliomas: Comparisonwith relative blood volume and tumor grade. AJNR Am J Neuroradiol. 2002;23:7-14.2. Provenzale JM, Wang GR, Brenner T, et al. Compar-ison of permeability in high-grade and low-grade brain tumors using dynamic susceptibility contrast MR imaging. AJR Am J Roentgenol. 2002;178:711–716.3. Covarrubias DJ, Rosen BR, Lev MH. Dynamic mag-netic resonance perfusion imaging of brain tumors.Oncologist. 2004;9:528-537.4. Lev MH, Rosen BR. Clinical applications of intracra-nial perfusion MR imaging. Neuroimaging Clin N Am.1999;9:309-331.5. Roberts HC, Roberts TP, Brasch RC, Dillon WP.Quantitative measurement of microvascular permeabil-ity in human brain tumors achieved using dynamic contrast-enhanced MR imaging: Correlation with histo-logic grade. AJNR Am J Neuroradiol. 2000;21:891-899.6. Henry RG, Vigneron DB, Fischbein NJ, et al. Com-parison of relative cerebral blood volume and protonspectroscopy in patients with treated gliomas. AJNRAm J Neuroradiol. 2000;21:357-366.7. Aronen HJ, Perkio J. Dynamic susceptibility contrastMRI of gliomas. Neuroimaging Clin N Am. 2002;12:501-523.8. Essig M, Waschkies M, Wenz F, et al. Assessment ofbrain metastases with dynamic susceptibility-weightedcontrast-enhanced MR imaging: Initial results. Radiol-ogy. 2003;228:193-199.9. Howe FA, Opstad KS. 1H MR spectroscopy ofbrain tumours and masses. NMR Biomed. 2003;16:123-131.10. Tzika AA, Zarifi MK, Goumnerova L, et al. Neu-roimaging in pediatric brain tumors: Gd-DTPA-enhanced, hemodynamic, and diffusion MR imagingcompared with MR spectroscopic imaging. AJNR Am J Neuroradiol. 2002;23:233-333.11. Castillo M, Kwock L, Mukherji SK. Clinical applica-tions of proton MR spectroscopy. AJNR Am J Neuro-radiol. 1996;17(1):1-15.12. Usenius JP, Kauppinen RA, Vaino PA, et al. Quanti-tative metabolite patterns of human brain tumors:Detection by 1H NMR spectroscopy in vivo and in vitro.J Comput Assist Tomogr. 1994;18:705-713.

BRAIN TUMOR MRI

18 � APPLIED RADIOLOGY© www.appliedradiology.com May 2006

13. Usenius JP, Vainio P, Hernesniemi J, et al.Choline-containing compounds in human astrocy-tomas studied by 1H NMR spectroscopy in vivo and invitro. J Neurochem. 1994;63:1538-1543.14. Nelson SJ. Multivoxel magnetic resonance spec-troscopy of brain tumors. Mol Cancer Ther. 2003;2:497-507.15. Law M, Cha S, Knopp EA, et al. High-grade glio-mas and solitary metastases: Differentiation by usingperfusion and proton spectroscopic MR imaging. Radi-ology. 2002;222:715-721.16. Howe FA, Barton SJ, Cudlip SA, et al. Metabolicprofiles of human brain tumors using quantitative invivo 1H magnetic resonance spectroscopy. MagnReson Med. 2003;49:223-232.17. Gill SS, Thomas DG, Van Bruggen N, et al. ProtonMR spectroscopy of intracranial tumours: In vivo and in vitro studies. J Comput Assist Tomogr. 1990;14:497-504. 18. Shino A, Nakasu S, Matsuda M, et al. Noninva-sive evaluation for the malignant potential of intracra-nial meningiomas performed using proton magneticresonance spectroscopy. J Neurosurg. 1999;91:928-934.19. Graves EE, Nelson SJ, Vigneron DB, et al. Serialproton MR spectroscopic imaging of recurrent malig-nant gliomas after gamma knife radiosurgery. AJNRAm J Neuroradiol. 2001;22:613-624.20. Vlieger EJ, Majoie CB, Leenstra S, Den HeetenGJ. Functional magnetic resonance imaging for neuro-surgical planning in neuroncology. Eur Radiol. 2004;14:1143-1153.21. Wilkinson ID, Romanowski CA, Jellinek DA, et al.Motor functional MRI for pre-operative and intraopera-tive neurosurgical guidance. Br J Radiol. 2003;76:98-103.22. Moritz C, Haughton V. Functional MR imaging:Paradigms for clinical preoperative mapping. MagnReson Imaging Clin N Am. 2003;11:529-542.

23. Huang J, Carr TH, Cao Y. Comparing cortical acti-vations for silent and overt speech using event-relatedfMRI. Hum Brain Mapp. 2002;15(1):39-53.24.Larner AJ, Robinson G, Kartsounis LD, et al. Clini-cal-anatomical correlation in a selective speech pro-duction impairment. J Neurol Sci. 2004;219:23-29.25. Li P, Jin Z, Tan LH. Neural representations ofnouns and verbs in Chinese: An fMRI study. Neuroim-age. 2004;21:1533-1541.26. Kho KH, Leijten FS, Rutten GJ, et al. Discrepantfindings for Wada test and functional magnetic reso-nance imaging with regard to language function: Useof electrocortical stimulation mapping to confirmresults. Case report. J Neurosurg. 2005;102:169-173.27. Schulze-Bonhage A, Quiske A, Loddenkemper T,et al. Validity of language lateralisation by unilateralintracarotid Wada test. J Neurol Neurosurg Psychiatry.2004;75:1367-1368.28. Petrovich NM, Holodny AI, Brennan CW, Gutin PH.Isolated translocation of Wernicke’s area to the righthemisphere in a 62-year-man with a temporo-parietalglioma. AJNR Am J Neuroradiol. 2004;25:130-133.29. Holodny AI, Schulder M, Ybasco A, Liu WC.Translocation of Broca’s area to the contralateral hemi-sphere as the result of the growth of a left inferiorfrontal glioma. J Comput Assist Tomogr. 2002;26:941-943.30. Derakhshan I. Unsuspected atypical hemisphericdominance for language as determined by FMRI.Handedness: Neural versus behavioral: The differ-ence is measurable. Epilepsia. 2003;44:734-735.31. Price CJ, Veltman DJ, Ashburner J, et al. The criti-cal relationship between the timing of stimulus presen-tation and data acquisition in blocked designs withfMRI. Neuroimage. 1999;10(1):36-44.32. Golby AJ, Poldrack RA, Brewer JB, et al. Material-specific lateralization in the medial temporal lobe andprefrontal cortex during memory encoding. Brain.2001;124(Pt 9):1841-1854.

33. Rutten GJ, Ramsey N, Noordmans HJ, et al.Toward functional neuronavigation: Implementation offunctional magnetic resonance imaging data in a surgi-cal guidance system for intraoperative identification ofmotor and language cortices. Technical note and illus-trative case. Neurosurg Focus. 2003;15(1):E6.34. Bogomolny DL, Petrovich NM, Hou BL, et al. Func-tional MRI in the brain tumor patient. Top Magn ResonImaging. 2004;15:325-335.35. Haberg A, Kvistad KA, Unsgard G, Haraldseth O.Preoperative blood oxygen level-dependent functionalmagnetic resonance imaging in patients with primarybrain tumors: Clinical application and outcome. Neuro-surgery. 2004;54:902-914.36. Nimsky C, Ganslandt O, Hastreiter P, et al. Intraop-erative diffusion-tensor MR imaging: Shifting of whitematter tracts during neurosurgical procedures—initialexperience. Radiology. 2005; 234:218-225.37. Sinha S, Bastin ME, Whittle IR, Wardlaw JM. Diffu-sion tensor MR imaging of high-grade cerebralgliomas. AJNR Am J Neuroradiol. 2002;23:520-527.38. Basser PJ, Jones DK. Diffusion-tensor MRI: The-ory, experimental design and data analysis—a techni-cal review. NMR Biomed. 2002;15:456-467.39. Le Bihan D, Mangin JF, Poupon C, et al. Diffusiontensor imaging: Concepts and applications. J MagnReson Imaging. 2001;13:534-546.40. Tropine A, Vucurevic G, Delani P, et al. Contribu-tion of diffusion tensor imaging to delineation ofgliomas and glioblastomas. J Magn Reson Imaging.2004;20:905-912.41. Holodny AI, Gor DM, Watts R, et al. Diffusion-ten-sor MR tractography of somatotopic organization ofcorticospinal tracts in the internal capsule: Initialanatomic results in contradistinction to prior reports.Radiology. 2005;234:649-653. 42. Conturo TE, Lori NF, Cull TS, et al. Tracking neu-ronal fiber pathways in the living human brain. ProcNatl Acad Sci USA. 1999;96:10422-10427.

BRAIN TUMOR MRI

Coming soonNormal Variants and Pitfalls in Whole Body PET Imaging with F-18 FDGVal J. Lowe, MD, Mayo Clinic, Rochester, MN; Trond Velde Bogsrud, MD, University Clinic,The Norwegian Radium Hospital, Oslo, Norway

MRI Determination of Myocardial ViabilityDavid A. Bluemke, MD, PhD; David Grand, MD, Johns Hopkins University School of Medicine, Baltimore, MD

AAA Stent-Grafting in 2006: What Have 10 Years Taught Us? Kenneth R. Thomson, MD, FRANZCR and Jeff D.C. Tam, MBBS, The Alfred, Melbourne,Australia; Peter Y. Milne, MB, BS, FRACS, FRCS (Eng), FACS, The Royal Melbourne Hospital, Melbourne, Australia; Tom McIntyre, MIR, William A. Cook, Australia Ltd., Brisbane, Australia

CT and MRI of Adrenal MassesAntonio Carlos A. Westphalen, MD and Bonnie N. Joe, MD, PhD, University of California,San Francisco, CA

MRI of the Ankle and Foot Part 1Muhammad Ali, MBBS and Tim S. Chen, MD, Radnet Management Inc., Los Angeles, CA;John V. Crues III, MD, Radnet Management Inc. and University of California, San Diego, CA