Simultaneous MR/PET Imagingof the Human Brain: FeasibilityStudy1

Heinz-Peter W. Schlemmer, MD, PhDBernd J. Pichler, MDMatthias Schmand, PhDZiad Burbar, MDChristian Michel, MDRalf Ladebeck, MDKirstin Jattke, MDDavid Townsend, PhDClaude Nahmias, MDPradeep K. Jacob, MDWolf-Dieter Heiss, MDClaus D. Claussen, MD

The purpose of this study was to apply a magnetic reso-nance (MR) imaging–compatible positron emission tomo-graphic (PET) detector technology for simultaneous MR/PET imaging of the human brain and skull base. The PETdetector ring consists of lutetium oxyorthosilicate (LSO)scintillation crystals in combination with avalanche photo-diodes (APDs) mounted in a clinical 3-T MR imager withuse of the birdcage transmit/receive head coil. Followingphantom studies, two patients were simultaneously exam-ined by using fluorine 18 fluorodeoxyglucose (FDG) PETand MR imaging and spectroscopy. MR/PET data enabledaccurate coregistration of morphologic and multifunctionalinformation. Simultaneous MR/PET imaging is feasible inhumans, opening up new possibilities for the emergingfield of molecular imaging.

� RSNA, 2008

1 From the Department of Diagnostic Radiology, UniversityHospital Tubingen, Hoppe-Seyler-Strasse 3, D-72076Tubingen, Germany (H.W.S., B.J.P., C.D.C.); Departmentof Molecular Imaging, Siemens Medical Solutions, Knox-ville, Tenn (M.S., Z.B., C.M.); Department of MagneticResonance, Siemens Medical Solutions, Erlangen, Ger-many (R.L., K.J.); Molecular Imaging and TranslationalResearch Program, University of Tennessee GraduateSchool of Medicine, Knoxville, Tenn (D.T., C.N.); Depart-ment of Radiology, University of Tennessee Medical Cen-ter, Knoxville, Tenn (P.K.J.); and Max Planck Institute forNeurological Research, Cologne, Germany (W.D.H.). Re-ceived November 12, 2007; revision requested January23, 2008; revision received March 25; accepted April 21;final version accepted April 24. Address correspondenceto H.P.S. (e-mail: heinz-peter.schlemmer@med.uni-tuebingen.de).

� RSNA, 2008

ORIG

INAL

RESE

ARCH

�TE

CHNI

CAL

DEVE

LOPM

ENTS

1028 Radiology: Volume 248: Number 3—September 2008

Note: This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, use the Radiology Reprints form at the end of this article.

The comprehensive assessment ofanatomic, functional and molecu-lar information is essential for mo-

lecular imaging. However, exact spatialcorrelation of imaging data acquired se-quentially with separate scanners is lim-ited, essentially because patient reposi-tioning causes differing section orienta-tions, as well as variations in organshape and position. Several approachesfor sophisticated image fusion by usingaffine and deformable transformationshave been developed but combining im-age information by using visual compar-ison of images positioned side by side isstill a commonly used approach in clini-cal practice. The most accurate dataalignment can nevertheless be achievedby using hybrid systems enabling tem-poral and spatial coregistration of mor-phologic and functional data in a singleexamination and without repositioningthe patient. A milestone was achieved in1998 when the first integrated positronemission tomography (PET)/computedtomography (CT) scanner was intro-duced, combining functional PET withanatomic CT imaging in a single exami-nation (1). Subsequently, PET/CT hasbeen proved a powerful diagnostic toolin oncology for imaging a number ofdifferent tumors (eg, lung and head-and-neck tumors [2]), although PETand CT are still performed sequentially.

Magnetic resonance (MR) imaginghas the distinct advantage, as comparedwith CT, of providing higher soft-tissuecontrast that yields superior diagnosticaccuracy, particularly in the brain, ab-dominal organs, and bone marrow(3,4). Particularly for brain imaging, si-multaneous MR and PET imaging wouldenable a precise combination of mor-phologic, functional, and metabolic in-formation for studying such processesas perfusion-dependent kinetics oftracer distribution. However, from atechnologic point of view, simultaneous

imaging was not feasible until now be-cause conventional PET detectors arehighly sensitive, even to small magneticfields (5). However, new PET detectortechnology that is unaffected by the highmagnetic field strength of the MR imag-ing system has been developed with lu-tetium oxyorthosilicate (LSO) scintilla-tion crystals and avalanche photodiodes(APDs) (6,7). Recently, it has beenshown that quantitative simultaneousMR/PET imaging is feasible in small an-imal systems (8).

The goal of this study was to demon-strate the feasibility of simultaneous MR/PET imaging of the human brain by usingthis PET detector technology integratedin a clinical 3.0-T MR imaging system.

Materials and Methods

PatientsInstitutional Review Board approval wasobtained for MR/PET examinations of pa-tients with a history of cancer who had un-dergone routine clinical fluorine 18 fluoro-deoxyglucose (FDG) PET/CT imaging forfollow-up. Informed consent was obtainedfrom each patient before the MR/PET ex-amination, which immediately followed thePET/CT examination. In this study, two pa-tients were examined: one with a history ofmalignant skin cancer and mild cognitiveimpairment and one with a history of thy-roid cancer. The MR/PET examinationswere performed in November 2006 (Sie-

mens Medical Solutions, Knoxville, Tenn).The hardware and software components ofthe system were provided by Siemens(BrainPET; Erlangen, Germany). Severalauthors (M.S., Z.B., C.M., R.L., and K.J.)are employees of Siemens Medical Solu-tions (Knoxville, Tenn) and one author(W.D.H.) is a consultant for Siemens Med-ical Solutions (Knoxville, Tenn). All authorshad control of the image data and postpro-cessing procedures, as well as all informa-tion used for the publication.

PET DetectorThe PET detector technology has beendeveloped by using LSO scintillation crys-tals (7) and APDs (6,7) that can be oper-ated within a high-field-strength MR im-aging system. Each block detector con-sists of a 12 � 12 LSO matrix with anindividual crystal size of 2.5 � 2.5 � 20mm. The block is displayed on a 3 � 3APD array with individual diodes thathave an active surface of 5 � 5 mm. SixLSO-APD block detectors form a cassetteand are arranged on the z axis of thescanner, forming 72 crystal rings. The en-tire PET scanner consists of 32 radiallyarranged cassettes and has an axial fieldof view (FOV) of about 19.2 cm and aninner ring diameter of 35.5 cm. The sys-tem has a time resolution of 5.6 msec (fullwidth at half maximum) and a mean en-ergy resolution at 511 keV of 22% (full

Published online10.1148/radiol.2483071927

Radiology 2008; 248:1028–1035

Abbreviations:APD � avalanche photodiodesFDG � fluorine 18 fluorodeoxyglucoseFOV � field of viewHRRT � high-resolution research tomographyLSO � lutetium oxyorthosilicate

Author contributions:Guarantors of integrity of entire study, H.W.S., D.T.,C.D.C.; study concepts/study design or data acquisition ordata analysis/interpretation, all authors; manuscript draft-ing or manuscript revision for important intellectual con-tent, all authors; approval of final version of submittedmanuscript, all authors; literature research, H.W.S.,B.J.P., M.S., C.N., W.D.H.; clinical studies, H.W.S., D.T.,C.N., P.K.J., W.D.H.; experimental studies, B.J.P., M.S.,Z.B., C.M., R.L., K.J.; and manuscript editing, H.W.S.,B.J.P., M.S., C.M., R.L., K.J., D.T., C.N., W.D.H., C.D.C.

See Materials and Methods for pertinent disclosures.

Advance in Knowledge

� This technology enables simulta-neous and isocentric MR/PET im-aging of the human brain with im-age quality comparable to stand-alone systems.

Implications for Patient Care

� The precise anatomic matchingachieved enables the investigationof metabolism and function insmall brain areas such as the nu-clei of basal brain and brainstem.

� Simultaneous PET and functionalMR imaging activation studiescombined with diffusion tensortracking may provide a deeperunderstanding of brain function.

� The effect of transmitter/receptoractivation on perfusion may beanalyzed by examining dynamicand perfusion-dependent kineticsof tracer/drug distribution in vari-ous brain structures.

TECHNICAL DEVELOPMENTS: Simultaneous MR/PET Imaging of the Human Brain Schlemmer et al

Radiology: Volume 248: Number 3—September 2008 1029

width at half maximum). All PET detectorcomponents such as amplifiers, resistors,shielding material, and housing were se-lected to minimize interference with theMR imaging B field, gradients, and radio-frequency. Large, solid, conductive mate-rials were avoided to reduce the inductionof eddy currents.

MR ImagerThe MR-compatible PET detector ringwas constructed to fit inside a standardclinical 3-T MR imager (TRIO; SiemensMedical Solutions) to match the twoFOVs. Furthermore, a standard bird-cage transmit/receive head coil (Sie-mens Medical Solutions) mounted onthe MR imager bed could be placed in-side the PET detector ring, therebymatching the two FOVs and enablingtemporal and spatial registration of PETand MR data (Fig 1).

Combined MR/PET Imaging and PhantomStudiesInitial evaluation of the PET detectorperformance and of potential mutual in-terference between the PET and MRsystems was undertaken by using a 20-cm-diameter cylindrical phantom filledwith 37 MBq of FDG.

A Hoffman brain phantom, filled with37 MBq of FDG, or alternatively, a high-resolution Cologne phantom (9,10) wasplaced in the MR/PET system. PET datawere acquired for 30 minutes while MRimaging was performed by using a T2-weighted turbo spin-echo sequence. Po-tential artifacts caused by mutual inter-ference between the PET and MR imag-ing system were qualitatively tested onthe PET detector, the MR imaging signallevel, and on image quality.

MR Sequence ProtocolThe MR examination was performedduring PET data acquisition. Conven-tional MR imaging of the brain includedan axial T2-weighted turbo spin-echosequence (repetition time msec/echotime msec, 4000/105; FOV, 216 � 240mm; matrix size, 288 � 320; voxel size,0.8 � 0.8 � 5.0 mm; and acquisitiontime, 2 minutes 8 seconds), a fluid-at-tenuated inversion recovery sequence(10 000/106; inversion time msec, 2500;

FOV, 180 � 240 mm; matrix size, 168 �320; voxel size, 1.1 � 0.8 � 5.0 mm; andacquisition time, 4 minutes 22 seconds),and a three-dimensional T1-weighted fastlow-angle shot sequence (23/7.4; flip an-gle, 25°; FOV, 176 � 256 mm; matrixsize, 176 � 256; voxel size, 1.0 � 1.0 �1.0 mm; and acquisition time, 6 minutes11 seconds). For diffusion-weighted im-aging, a single-shot echo-planar imagingsequence was applied (3000/791; FOV,240 � 240 mm; matrix size, 128 � 128;voxel size, 1.9 � 1.9 � 5.0 mm; acquisi-tion time, 1 minute 11 seconds). Waterdiffusion was measured with a three-image trace technique by using b values of0, 500, and 1000 sec/mm2. Time-of-flightMR angiography was conducted at thelevel of the circle of Willis by using athree-dimensional fast low-angle shot se-quence (22/3.7; flip angle, 18°; FOV,150 � 200 mm; matrix size, 202 � 384;voxel size, 0.7 � 0.5 � 06 mm; and ac-quisition time, 4 minutes 19 seconds).For proton MR spectroscopy, two-dimen-sional spin-echo chemical shift imagingsequences were applied with a matrix sizeof 10 � 10 (point-resolved spectroscopysequence for localization, elliptic k-spacesampling). For patient 1, an echo time of135 msec was chosen and the sequencewas performed twice, in an axial orienta-

tion at the level of the ventricles and in acoronal orientation at the level of the hip-pocampus (1700/135; voxel size, 1.0 �1.0 � 1.5 mm; number of excitations, 3;and acquisition time, 6 minutes 46 sec-onds). In patient 2, the imaging was per-formed on an axial orientation at the levelof the ventricles (1700/30; voxel size,0.75 � 1.0 � 1.5 mm; number of signalsacquired, three; and acquisition time, 4minutes 51 seconds). MR data were qual-itatively evaluated regarding line broad-ening of resonances in proton MR spec-tra, as well as overall quality and potentialartifacts of MR images.

PET ImagingFor PET imaging, the patients fasted forat least 4 hours before the intravenousinjection of 370 MBq of FDG. The distri-bution of the tracer was recorded for 20or 40 minutes at a steady state, 120 min-utes after injection for the PET/CT scan atthe University of Tennessee Medical Cen-ter (Knoxville, Tenn). Repeated multiplesamples of arterial blood could not be ob-tained in the setting of this conceptualstudy and therefore, quantification ofmetabolic rates for glucose was not at-tempted. Data were acquired in 96-bit listmode containing complete informationfor each coincident event and were stored

Figure 1

Figure 1: Drawing and photograph of integrated MR/PET design shows isocentric layering of MR headcoil, PET detector ring, and tunnel of MR magnet.

TECHNICAL DEVELOPMENTS: Simultaneous MR/PET Imaging of the Human Brain Schlemmer et al

1030 Radiology: Volume 248: Number 3—September 2008

on a disk, accumulating approximately 20GB of raw data during the 40 minutes ofacquisition. Prompt and random coinci-dences were then rebinned in sinogramsin a high-resolution mode (span, 9; ringdifference, 67), resulting in a three-dimensional data set of 1399 sinograms,each consisting of 256 angular projectionsand 192 radial elements. The data set wasreconstructed iteratively (six iterationsand 16 subsets) on a computer (approxi-mately 12 seconds per iteration) by usingan ordinary Poisson ordered subset ex-pectation maximization three-dimen-sional (OPOSEM3D) algorithm (11) fromprompt coincidences, normalization, at-tenuation, and expected random (12) andscatter (13) coincidences. The image vol-ume consists of 153 transaxial imageplanes with a voxel size of 1.25 mm3. Re-constructed resolution was 2.5 mm fullwidth at half maximum in the center and4.5 mm at 10 cm off axis (for a line sourcein air reconstructed by using OPOSEM3D).For attenuation correction, the image ofthe head obtained by using MR was seg-mented in air and tissue (bone was ne-glected). Labels were assigned to eachregion and were then converted to atten-uation coefficients at 511 keV and for-ward projected. The normalization wascalculated from four 8-hour plane sourcenormal images. The plane source extend-ing the cross section of the entire FOVwas oriented in the z axis and rotated fourtimes with a 36° step angle. The activityimages of the volume of the brain weredisplayed in an orthogonal navigator(transaxial, sagittal, and frontal views) asmultisection (volume) images. The VINCI(volume-imaging in neurological re-search, coregistration and ROIs included[14]) image fusion software allowed lin-ear transformation and rotation to over-lay the PET and MR images. This imagefusion will be done in the future automat-ically by using a predetermined transfor-mation matrix.

Results

PET Detector and MR ImagerThe registration of annihilation photonsby the scintillation crystals was not no-ticeably compromised either by the

static magnetic field or by the simulta-neous switching of radiofrequency andmagnetic gradients. Flood images ac-quired with the PET detectors insideand outside MR imaging and while ac-quiring gradient-echo and spin-echo se-quences showed no obvious degrada-tion in the ability to identify the individ-ual crystals and their energy spectrawithin the detector blocks. Energy reso-lution remained constant for the blocksevaluated. Pickup noise seen on thebaseline level of the PET signals, cov-ered by gradient switching, was elimi-nated by setting the constant fractiondiscriminator at a lower energy level of250 keV. Count rate statistics were sim-ilar whether the PET insert was usedoutside or inside the MR imager. ThePET detector ring revealed a pointsource sensitivity of more than 6% withno MR coil or patient handling device inthe FOV of the PET scanner. The en-ergy window for the sensitivity mea-surement was set between 400 and 650keV. Cylindrical phantom measure-ments revealed homogeneity of betterthan 10% of the reconstructed PET im-age over a measured interval of 4 hours.For this measurement, all correctionswere applied, including calculated at-tenuation correction since the PETscanner was operated outside the MRsystem.

Combined MR/PET Imaging and PhantomStudiesThe performance of the MR imagingsystem was not qualitatively influ-enced by the integrated PET detectorring. Simultaneous MR/PET data ac-quisition from the Hoffman brainphantom revealed an MR and PET im-age quality comparable to those ofstand-alone systems with no evidenceof major artifacts or image distortions(Fig 2a). In a few central sections ofthe reconstructed PET images, minorring artifacts were observed owing tothe unoptimized state of the PET de-tector normalization, which was madeon the basis of a pseudo-rotating planesource (37 MBq of germanium 68). Byusing the high-resolution phantom, ra-dioactive rods of 3.0 mm in diametercould be separated at a contrast of

50% (Fig 2b). The Cologne phantomwas imaged with high-resolution re-search tomography (HRRT) (10,15)and shows comparable image resolu-tion, whereas the 2-mm structure canalmost be visually identified withHRRT (10). The homogeneity of MRimaging measured with an oil-filled24-cm homogeneous spheric phantomshowed degradations of less than 2parts per million when measurementswere obtained with the PET insert andcompared with those obtained withoutthe PET insert. No significant changein background noise was observed andthe signal-to-noise ratio remainedwithin the specifications of the MRsystem.

Combined MR/PET Imaging and PatientStudiesPET and MR image and spectroscopicquality was comparable to that of thecorresponding stand-alone scanners(Figs 3, 4). The PET images of FDGuptake showed the typical distributionof cerebral metabolic activity with highvalues in the cortex, basal ganglia, andthalamus; intermediate values in thecerebellum; and low values in the brain-stem and white matter. The spatial res-olution was similar to that of HRRT (12)and permitted the identification of smallnuclei in the basal brain and in thebrainstem (13). The patient who wasscanned for restaging of thyroid cancerrevealed a normal FDG uptake patternin the brain. In contrast, the patientwith mild cognitive impairment showedsome cortical areas and basal gangliaregions with moderately decreasedFDG uptake that corresponded to smallwhite matter hyperintensities detectedon T2-weighted MR images with in-creased choline-to-creatine signal inten-sity ratios on the MR spectroscopic im-ages (Fig 4).

Discussion

The results from phantom and humanstudies demonstrated that simultaneoushigh-field-strength MR and PET imagingof the human brain is feasible and theoverall performance of each modality ispreserved. The integrated LSO-APD

TECHNICAL DEVELOPMENTS: Simultaneous MR/PET Imaging of the Human Brain Schlemmer et al

Radiology: Volume 248: Number 3—September 2008 1031

Figure 2

Figure 2: (a) PET images of Hoffman brain phantom filled with 37 MBq of FDG. Images acquired with MR/PET system demonstrate quality comparable with stand-alone systems with no evidence of artifacts or distortions. (b) Cologne phantom– hot rod image shows 3-mm structures (horizontal, central line) that can clearly be sepa-rated with 50% contrast. Center spacing between hot rods is twice diameter of hole.

Figure 3

Figure 3: Simultaneously acquired (a) MR, (b) PET, and (c) superimposed combined MR/PET images of 66-year-old man after intravenous injection of 370 MBq ofFDG. Tracer distribution was recorded for 20 minutes at steady state after 120 minutes.

TECHNICAL DEVELOPMENTS: Simultaneous MR/PET Imaging of the Human Brain Schlemmer et al

1032 Radiology: Volume 248: Number 3—September 2008

PET detector did not affect the opera-tion of the MR imaging system, and like-wise, the PET detector performancewas not compromised during MR imag-ing. The MR/PET data that were simul-taneously acquired demonstrated animage quality comparable to stand-alone systems without any major distor-tions or artifacts. Temporal and spatialintrinsic registration of high-resolutionmorphologic and functional informationin different brain regions enabled accu-rate matching of FDG uptake and MRsignal characteristics. The new MR/PET–inserted PET scanner has a pointsource sensitivity of more than 6%(PET-only mode). The high sensitivity ofthe PET scanner and the ability of thecombined scanner to acquire multifunc-tion information from PET and MR im-aging (eg, functional MR imaging, perfu-sion- and diffusion-weighted MR imag-ing, MR spectroscopy) (Fig 4) make theMR/PET system an ideal device for ad-vanced studies of the human brain.

The combined system has severaladvantages when compared with equiv-alent freestanding systems: (a) The ra-diation dose is equivalent to stand-aloneMR and PET systems because the pa-tient is not exposed to additional ioniz-ing radiation. Moreover, we expect thatthe PET insert achieves sensitivity thatis close to that of brain HRRT (16).Therefore, it can be expected thatthe overall injected tracer dose might bereduced in the future, minimizing theradiation dose for the patient. (b) Thescanning time is reduced because of si-multaneous data acquisition. (c) Thequalitative MR image evaluation yieldedno obvious image degradation or arti-facts compared with conventional brainMR images obtained at 3.0 T. (d) Noloss of PET image quality was observedwhen comparing phantom measure-ments inside and outside the magnet.However, no National Electrical Manu-facturers Association measurementswere performed at this point. Doyle andring artifacts that can be seen in thePET images are not related to operationinside the magnet. Rather, they werecaused by the unfavorable ratio of thePET system ring diameter and transax-ial detector block gaps, required for the

geometry of the dedicated brain scan-ner. Similar effects have been reportedfor the HRRT brain scanner (17). Weare currently working on an improveddata normalization procedure. (e) Allapplied MR sequences, including diffu-sion-weighted imaging and MR spec-troscopy, are standard sequences pro-vided by the manufacturer. However, atthis stage of engineering, only the trans-mit/receive head coil has been imple-mented. Because the PET insert isshielded against electronic pick-up fromthe radiofrequency coils, further trans-mit/receive coils that can fit in the PETscanner are, in principle, feasible (eg, awrist coil).

Advantages of simultaneous data ac-quisition include temporal correlation ofPET and MR imaging data, reduced im-aging time compared with sequentialimaging, and improved spatial correla-tion owing to reduced motion artifacts.Furthermore, cardiac and respiratorymotion correction of PET data can beimplemented given MR imaging infor-mation. Technical challenges of the PETdetector system were related to the re-quired MR compatibility, including in-sensitivity to the high-strength magneticfield and the narrow diameter of thebore of the magnet. The magnetic field–and space-related constraints requireda modification of the PET detector tech-

Figure 4

Figure 4: Complete set of acquired MR data of same patient as in Figure 3 includes following imagingsequences: (a) T2-weighted turbo spin-echo, (b) fluid-attenuated inversion recovery (FLAIR), (c) time-of-flight MR angiography. (Fig 4 continues.)

TECHNICAL DEVELOPMENTS: Simultaneous MR/PET Imaging of the Human Brain Schlemmer et al

Radiology: Volume 248: Number 3—September 2008 1033

nology and geometry, which affectedimage normalization, the cooling sys-tem, and the ability to perform conven-tional transmission scans for attenua-tion correction. Accordingly, new meth-ods for MR imaging– based PETattenuation correction have to be devel-oped.

An inherent limitation of PET is thatit usually maps only one metabolic, mo-lecular, or functional parameter at atime. Therefore, the metabolic, molecu-lar, or functional objective for PET hasto be carefully determined and thetracer must be selected accordingly.However, by using the MR/PET device,several MR-derived morphologic andfunctional parameters can be observedconcurrently with PET imaging. Fur-thermore, as compared with sequen-tially acquired MR and PET imaging, thetime for data acquisition is notably re-duced.

Future applications of MR/PET willinclude simultaneous acquisition of PETand functional MR imaging combinedwith diffusion tensor tracking duringbrain activation studies. In addition,metabolic activity in small cortical areasand in small nuclei of the basal brainand brainstem may be studied by com-bining the exactly matched functionalPET images with the morphologic MRimages that are distinguished by usinghigh spatial and contrast resolution. Ki-netic studies may include the investiga-tion of dynamic distributions of tracersor drugs in various brain structures andtheir flow-dependent kinetics, as well asthe effects of transmitter/receptor acti-vation on perfusion. Moreover, the de-velopment of cell therapy may be moni-tored (eg, in Parkinson disease). Thetechnique has considerable potentialbut cannot be fully assessed at this earlystage. In any case, the technology willopen up a powerful window for brainresearch and for studies of brain dis-ease.

Limitations of this study includedthe lack of quantitative data from de-tailed investigations of mutual interfer-ence between the PET and MR imagingfrom phantom and additional humanstudies. Further detailed tests that useNational Electrical Manufacturers Asso-

Figure 4 (continued)

Figure 4 (continued): (d) diffusion-weighted MR, and (e) proton MR spectroscopic imaging. Increasedsignal intensities of resonances of choline relative to creatine are apparent in areas of white matter hyperinten-sities seen on FLAIR images (left spectra) compared with normal gray matter (right spectra). Cho � choline,Cr � creatine, NAA � N-acetylaspartate.

TECHNICAL DEVELOPMENTS: Simultaneous MR/PET Imaging of the Human Brain Schlemmer et al

1034 Radiology: Volume 248: Number 3—September 2008

ciation measurements and patient ex-aminations are currently planned butare beyond the scope of this initial feasi-bility report. To bring MR/PET to astage of accuracy where PET/CT cur-rently is will require considerably moreresearch and technical development.

This proof-of-concept study is fo-cused on brain imaging. Ongoing re-search in whole-body imaging, particu-larly with regard to oncology, will affordthe opportunity for comprehensive as-sessment of morphology, perfusion,function, and metabolism in other partsof the body. However, it is unlikely thatMR/PET will replace PET/CT. An im-portant advantage of PET/CT is that CTinformation can be used for PET attenu-ation correction. The attenuation cor-rection for PET on the basis of MR im-aging information is still challenging, al-though some early approaches haveproduced promising results (18). Nev-ertheless, substantial scientific and clin-ical progress can be anticipated by usingand advancing this MR/PET technology.

In conclusion, simultaneous high-field-strength MR and PET imaging ofthe human brain is feasible and theoverall performance of each modality ispreserved.

References1. Beyer T, Townsend DW, Brun T, et al. A

combined PET/CT scanner for clinical oncol-ogy. J Nucl Med 2000;41(8):1369–1379.

2. Goerres GW, von Schulthess GK, SteinertHC. Why most PET of lung and head-and-neck cancer will be PET/CT. J Nucl Med2004;45(suppl 1):66S–71S.

3. Muller-Horvat C, Radny P, Eigentler TK,et al. Prospective comparison of the impact

on treatment decisions of whole-body mag-netic resonance imaging and computed to-mography in patients with metastatic malig-nant melanoma. Eur J Cancer 2006;42(3):342–350.

4. Pfannenberg C, Aschoff P, Schanz S, et al.Prospective comparison of 18F-fluorodeoxy-glucose positron emission tomography/com-puted tomography and whole-body magneticresonance imaging in staging of advancedmalignant melanoma. Eur J Cancer 2007;43(3):557–564.

5. Pichler BJ, Judenhofer MS, Catana C, et al.Performance test of an LSO-APD detector ina 7-T MRI scanner for simultaneous PET/MRI. J Nucl Med 2006;47(4):639–647.

6. Melcher CL, Schweitzer JS. Cerium-dopedlutetium oxyorthosilicate: a fast, efficientnew scintillator. IEEE Trans Nucl Sci 1992;39(4):502–505.

7. Pichler B, Lorenz E, Mirzoyan R, et al. Per-formance test of a LSO-APD PET module in a9.4 Tesla magnet. In: Conference Record ofthe 1998 Institute of Electrical and Electron-ics Engineers (IEEE) Nuclear Science Sym-posium and Medical Imaging Conference.Piscataway, NJ: IEEE, 1998;1237:–1239.

8. Judenhofer MS, Catana C, Swann BK, et al.PET/MR images acquired with a compactMR-compatible PET detector in a 7-T mag-net. Radiology 2007;244(3):807–814.

9. Hoffman EJ, Cutler PD, Digby WM, et al.3-D phantom to simulate cerebral blood flowand metabolic images for PET. IEEE TransNucl Sci 1990;37(2):616–620.

10. Schmand M, Wienhard K, Casey ME, et al.Performance evaluation of a new LSO highresolution researchtomograph-HRRT. In:Conference Record of the 1999 Institute ofElectrical and Electronics Engineers (IEEE)Nuclear Science Symposium and Medical Im-aging Conference. Piscataway, NJ: IEEE,1999;1067:–1071.

11. Comtat C, Bataille F, Michel C, et al.OSEM-3D reconstruction strategies for the

ECAT HRRT. In: Conference Record of the2004 Institute of Electrical and ElectronicsEngineers (IEEE) Nuclear Science Sympo-sium and Medical Imaging Conference. Pis-cataway, NJ: IEEE, 2004;3492:–3496.

12. Byars LG, Sibomana M, Burbar Z. Variancereduction on randoms from delayed coinci-dence histograms for the HRRT. In: Confer-ence Record of the 2005 Institute of Electri-cal and Electronics Engineers (IEEE) NuclearScience Symposium and Medical ImagingConference. Piscataway, NJ: IEEE, 2005;2622:–2626.

13. Watson CC, Casey ME, Michel C, BendriemB. Advances in scatter correction for 3DPET/CT. In: Conference Record of the 2004Institute of Electrical and Electronics Engi-neers (IEEE) Nuclear Science Symposiumand Medical Imaging Conference. Piscat-away, NJ: IEEE, 2004;3008:–3012.

14. Vollmar S, Hampl JA, Kracht L, et al. Inte-gration of functional data (PET) into brainsurgery planning and neuronavigation. In:Buzug TM, ed. Advances in medicalengineering: proceedings in physics 114.Berlin, Germany: Springer, 2007;98:–103.

15. Wienhard K, Schmand M, Casey ME, et al.The ECAT HRRT: performance and first clin-ical application of the new high resolutionresearch tomograph. IEEE Trans Nucl Sci2002;49(1):104–110.

16. Heiss WD, Habedank B, Klein JC, et al. Met-abolic rates in small brain nuclei determinedby high-resolution PET. J Nucl Med 2004;45(11):1811–1815.

17. Vollmar S, Michel C, Treffert JT, et al.HeinzelCluster: accelerated reconstructionfor FORE and OSEM3D. Phys Med Biol2002;47(15):2651–2658.

18. Hofmann M, Steinke F, Scheel V, et al. MR-based attenuation correction for PET/MR: anovel approach combining pattern recogni-tion and atlas registration. J Nucl Med 2008[in press].

TECHNICAL DEVELOPMENTS: Simultaneous MR/PET Imaging of the Human Brain Schlemmer et al

Radiology: Volume 248: Number 3—September 2008 1035

Radiology 2008 This is your reprint order form or pro forma invoice

(Please keep a copy of this document for your records.)

Reprint order forms and purchase orders or prepayments must be received 72 hours after receipt of form either by mail or by fax at 410-820-9765. It is the policy of Cadmus Reprints to issue one invoice per order.

Please print clearly.

Author Name _______________________________________________________________________________________________ Title of Article _______________________________________________________________________________________________ Issue of Journal_______________________________ Reprint # _____________ Publication Date ________________ Number of Pages_______________________________ KB # _____________ Symbol Radiology Color in Article? Yes / No (Please Circle) P lease include the journal name and reprint number or manuscript number on your purchase order or other correspondence.

Order and Shipping Information Reprint Costs (Please see page 2 of 2 for reprint costs/fees.) ________ Number of reprints ordered $_________ ________ Number of color reprints ordered $_________ ________ Number of covers ordered $_________ Subtotal $_________ Taxes $_________ (Add appropriate sales tax for Virginia, Maryland, Pennsylvania, and the District of Columbia or Canadian GST to the reprints if your order is to be shipped to these locations.) First address included, add $32 for each additional shipping address $_________

TOTAL $_________

Country _________________________________________

Quantity __________________ Fax __________________

Shipping Address (cannot ship to a P.O. Box) Please Print Clearly Name ___________________________________________ Institution _________________________________________ Street ___________________________________________ City ____________________ State _____ Zip ___________ Country ___________________________________________ Quantity___________________ Fax ___________________ Phone: Day _________________ Evening _______________ E-mail Address _____________________________________ Additional Shipping Address* (cannot ship to a P.O. Box) Name ___________________________________________ Institution _________________________________________ Street ___________________________________________ City ________________ State ______ Zip ___________

Phone: Day ________________ Evening ______________ E-mail Address ____________________________________ * Add $32 for each additional shipping address

P ayment and Credit Card Details Enclosed: Personal Check ___________ Credit Card Payment Details _________ Checks must be paid in U.S. dollars and drawn on a U.S. Bank. Credit Card: __ VISA __ Am. Exp. __ MasterCard Card Number __________________________________ E xpiration Date_________________________________ Signature: _____________________________________ Please send your order form and prepayment made payable to: Cadmus Reprints P.O. Box 751903 Charlotte, NC 28275-1903 Note: Do not send express packages to this location, PO Box.

FEIN #:541274108

Invoice or Credit Card Information Invoice Address Please Print Clearly Please complete Invoice address as it appears on credit card statement Name ____________________________________________ Institution ________________________________________ Department _______________________________________ Street ____________________________________________ City ________________________ State _____ Zip _______ Country ___________________________________________ Phone _____________________ Fax _________________ E-mail Address _____________________________________ Cadmus will process credit cards and Cadmus Journal

Services will appear on the credit card statement. If you don’t mail your order form, you may fax it to 410-820-9765 with

your credit card information.

Signature __________________________________________ Date _______________________________________ Signature is required. By signing this form, the author agrees to accept the responsibility for the payment of reprints and/or all charges described in this document. RB-9/26/07

Page 1 of 2

Black and White Reprint Prices

Domestic (USA only) # of

Pages 50 100 200 300 400 500

1-4 $221 $233 $268 $285 $303 $323 5-8 $355 $382 $432 $466 $510 $544 9-12 $466 $513 $595 $652 $714 $775

13-16 $576 $640 $749 $830 $912 $995 17-20 $694 $775 $906 $1,017 $1,117 $1,22021-24 $809 $906 $1,071 $1,200 $1,321 $1,47125-28 $928 $1,041 $1,242 $1,390 $1,544 $1,68829-32 $1,042 $1,178 $1,403 $1,568 $1,751 $1,924

Covers $97 $118 $215 $323 $442 $555

International (includes Canada and Mexico) # of

Pages 50 100 200 300 400 500

1-4 $272 $283 $340 $397 $446 $506 5-8 $428 $455 $576 $675 $784 $884 9-12 $580 $626 $805 $964 $1,115 $1,278

13-16 $724 $786 $1,023 $1,232 $1,445 $1,65217-20 $878 $958 $1,246 $1,520 $1,774 $2,03021-24 $1,022 $1,119 $1,474 $1,795 $2,108 $2,42625-28 $1,176 $1,291 $1,700 $2,070 $2,450 $2,81329-32 $1,316 $1,452 $1,936 $2,355 $2,784 $3,209

Covers $156 $176 $335 $525 $716 $905 Minimum order is 50 copies. For orders larger than 500 copies, please consult Cadmus Reprints at 800-407-9190. Reprint Cover Cover prices are listed above. The cover will include the publication title, article title, and author name in black. Shipping Shipping costs are included in the reprint prices. Domestic orders are shipped via UPS Ground service. Foreign orders are shipped via a proof of delivery air service. Multiple Shipments Orders can be shipped to more than one location. Please be aware that it will cost $32 for each additional location. Delivery Your order will be shipped within 2 weeks of the journal print date. Allow extra time for delivery.

Color Reprint Prices

Domestic (USA only) # of

Pages 50 100 200 300 400 500

1-4 $223 $239 $352 $473 $597 $719 5-8 $349 $401 $601 $849 $1,099 $1,3499-12 $486 $517 $852 $1,232 $1,609 $1,992

13-16 $615 $651 $1,105 $1,609 $2,117 $2,62417-20 $759 $787 $1,357 $1,997 $2,626 $3,26021-24 $897 $924 $1,611 $2,376 $3,135 $3,90525-28 $1,033 $1,071 $1,873 $2,757 $3,650 $4,53629-32 $1,175 $1,208 $2,122 $3,138 $4,162 $5,180

Covers $97 $118 $215 $323 $442 $555

International (includes Canada and Mexico)) # of

Pages 50 100 200 300 400 500

1-4 $278 $290 $424 $586 $741 $904 5-8 $429 $472 $746 $1,058 $1,374 $1,6909-12 $604 $629 $1,061 $1,545 $2,011 $2,494

13-16 $766 $797 $1,378 $2,013 $2,647 $3,28017-20 $945 $972 $1,698 $2,499 $3,282 $4,06921-24 $1,110 $1,139 $2,015 $2,970 $3,921 $4,87325-28 $1,290 $1,321 $2,333 $3,437 $4,556 $5,66129-32 $1,455 $1,482 $2,652 $3,924 $5,193 $6,462

Covers $156 $176 $335 $525 $716 $905 Tax Due Residents of Virginia, Maryland, Pennsylvania, and the District of Columbia are required to add the appropriate sales tax to each reprint order. For orders shipped to Canada, please add 7% Canadian GST unless exemption is claimed. Ordering Reprint order forms and purchase order or prepayment is required to process your order. Please reference journal name and reprint number or manuscript number on any correspondence. You may use the reverse side of this form as a proforma invoice. Please return your order form and prepayment to: Cadmus Reprints P.O. Box 751903 Charlotte, NC 28275-1903 Note: Do not send express packages to this location, PO Box. FEIN #:541274108

Reprint Order Forms and purchase order or prepayments must be received 72 hours after receipt of form.

Please direct all inquiries to:

Rose A. Baynard 800-407-9190 (toll free number) 410-819-3966 (direct number) 410-820-9765 (FAX number)

baynardr@cadmus.com (e-mail)

Page 2 of 2

Radiology 2008