5
UNCORRECTED PROOF Editorial Putting clearinto nuclear medicine: a decade of PET/CT development Thomas Beyer 1 , David W. Townsend 2 1 Department of Nuclear Medicine, University Hospital Essen, Hufelandstrasse 55, 45122 Essen, Germany 2 Departments of Medicine and Radiology, University of Tennessee Medical Center Knoxville, TN 37920 USA Published online: 23 June 2006 © Springer-Verlag 2006 Eur J Nucl Med Mol Imaging (2006) 33:857861 DOI 10.1007/s00259-006-0137-z Combining data from different imaging modalities is far from being a new idea. Hand-drawn contours on early thyroid scans constituted a rough anatomical reference for images of thyroid function acquired using trace amounts of radioactive iodine. When analogue imaging with planar X-rays and Polaroid photos of radiotracer distributions gave way to the digital imaging era of the CT scanner and emission tomography, visual correlation of structure and function became possible. The next step, beginning in the late 1980s, was the development of computer software to co-register the images from different modalities, a development that was applied primarily to brain imaging. However, even though non-invasive anatomical and functional imaging became well-established diagnostic procedures within radiology and nuclear medicine, little attention was actually paid to the co-registration of images from the different modalities. What were the reasons for this? One reason was that accurate registration of anatomy and function was never really considered essential for the interpretation of the images and access to multimodality data was rarely routine and certainly not straightforward. Furthermore, anatomical and functional images were typically acquired in different departments and read by different specialists, a situation that did little to promote the use of software fusion techniques. Thus, up until the mid 1990s, and despite the expenditure of considerable intellectual effort, image fusion techniques had little impact on public healthcare. The arrival of PET/CT has changed this situation forever. A PET/CT prototype To overcome the general lack of enthusiasm for the use of retrospective image fusion, a different approach was necessary that offered routine availability of co-registered data sets. Rather than co-register the images sets post hoc, an approach proposed in the mid 1990s was to fuse the technologies into a single device that imaged both anatomy and function such that the resultant image sets were intrinsically aligned [1]. Then, with dual-modality tech- nology available, physicians would be encouragedto routinely use fused images. The development of the first device [2] that combined positron emission tomography (PET) and computed tomography (CT) commenced in 1995, over a decade ago, entering clinical evaluation at the University of Pittsburgh Medical Center in 1998 [3] and imaging a total of around 300 cancer patients [4]. Interest among physicians was initially mixed; the intrinsically fused PET/CT images were more rapidly accepted by surgeons and oncologists, who found the superimposition of functional information on familiar CT scans to be very helpful. Radiologists and nuclear medicine clinicians, however, were rather more wary of the potential implica- tions of combined PET/CT imaging. Depending on the viewpoint, the PET/CT was considered to be either a nuclear medicine device with improved attenuation cor- rection or a radiological device with the capability to image a new contrast agentthe fluorinated analog of glucose, FDG. Nuclear medicine physicians wondered if they really needed all that detailed CT anatomy to interpret PET scans that they had for many years been reading clinically. Above all, of course, acquisition and control of the new imaging device became something of a contentious issue. In addition to assessing the clinical impact of the PET/ CT prototype, the daily operation presented a number of unique challenges. PET scanners are generally operated by nuclear medicine technologists and CT scanners by Thomas Beyer ()) Department of Nuclear Medicine, University Hospital Essen, Hufelandstrasse 55, 45122 Essen, Germany e-mail: [email protected] Fax: +41-43-3439072 European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 8, August 2006

Putting ‘clear’ into nuclear medicine: a decade of PET/CT development

Embed Size (px)

Citation preview

Page 1: Putting ‘clear’ into nuclear medicine: a decade of PET/CT development

UNCO

RREC

TEDPR

OOF

Editorial

Putting ‘clear’ into nuclear medicine: a decade of PET/CTdevelopmentThomas Beyer1, David W. Townsend2

1 Department of Nuclear Medicine, University Hospital Essen, Hufelandstrasse 55, 45122 Essen, Germany2 Departments of Medicine and Radiology, University of Tennessee Medical Center Knoxville, TN 37920 USA

Published online: 23 June 2006© Springer-Verlag 2006

Eur J Nucl Med Mol Imaging (2006) 33:857–861DOI 10.1007/s00259-006-0137-z

Combining data from different imaging modalities is farfrom being a new idea. Hand-drawn contours on earlythyroid scans constituted a rough anatomical reference forimages of thyroid function acquired using trace amounts ofradioactive iodine. When analogue imaging with planarX-rays and Polaroid photos of radiotracer distributionsgave way to the digital imaging era of the CT scanner andemission tomography, visual correlation of structure andfunction became possible. The next step, beginning inthe late 1980s, was the development of computer softwareto co-register the images from different modalities, adevelopment that was applied primarily to brain imaging.However, even though non-invasive anatomical andfunctional imaging became well-established diagnosticprocedures within radiology and nuclear medicine, littleattention was actually paid to the co-registration of imagesfrom the different modalities. What were the reasons forthis? One reason was that accurate registration of anatomyand function was never really considered essential for theinterpretation of the images and access to multimodalitydata was rarely routine and certainly not straightforward.Furthermore, anatomical and functional images weretypically acquired in different departments and read bydifferent specialists, a situation that did little to promotethe use of software fusion techniques. Thus, up until themid 1990s, and despite the expenditure of considerable

intellectual effort, image fusion techniques had littleimpact on public healthcare. The arrival of PET/CT haschanged this situation forever.

A PET/CT prototype

To overcome the general lack of enthusiasm for the use ofretrospective image fusion, a different approach wasnecessary that offered routine availability of co-registereddata sets. Rather than co-register the images sets post hoc,an approach proposed in the mid 1990s was to fuse thetechnologies into a single device that imaged both anatomyand function such that the resultant image sets wereintrinsically aligned [1]. Then, with dual-modality tech-nology available, physicians would be “encouraged” toroutinely use fused images. The development of the firstdevice [2] that combined positron emission tomography(PET) and computed tomography (CT) commenced in1995, over a decade ago, entering clinical evaluation at theUniversity of Pittsburgh Medical Center in 1998 [3] andimaging a total of around 300 cancer patients [4]. Interestamong physicians was initially mixed; the intrinsicallyfused PET/CT images were more rapidly accepted bysurgeons and oncologists, who found the superimpositionof functional information on familiar CT scans to be veryhelpful. Radiologists and nuclear medicine clinicians,however, were rather more wary of the potential implica-tions of combined PET/CT imaging. Depending on theviewpoint, the PET/CT was considered to be either anuclear medicine device with improved attenuation cor-rection or a radiological device with the capability to imagea new contrast agent—the fluorinated analog of glucose,FDG. Nuclear medicine physicians wondered if they reallyneeded all that detailed CT anatomy to interpret PET scansthat they had for many years been reading clinically. Aboveall, of course, acquisition and control of the new imagingdevice became something of a contentious issue.

In addition to assessing the clinical impact of the PET/CT prototype, the daily operation presented a number ofunique challenges. PET scanners are generally operated bynuclear medicine technologists and CT scanners by

Thomas Beyer ())Department of Nuclear Medicine,University Hospital Essen,Hufelandstrasse 55,45122 Essen, Germanye-mail: [email protected]: +41-43-3439072

European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 8, August 2006

Page 2: Putting ‘clear’ into nuclear medicine: a decade of PET/CT development

UNCO

RREC

TEDPR

OOF

radiology technologists. Who, therefore, should operate acombined device? Obviously, the ideal solution was atechnologist trained in both CT and PET; in 1998 inPittsburgh, dual-certified technologists were not available,and therefore the original prototype had to be operated bytwo technologists, one for CT and one for PET. Thisarrangement was inefficient, especially for the CT tech-nologist, who came from radiology simply to push twobuttons, leaving the PET technologist to acquire theremainder of the scan, which could take up to an hour.Processing of the data would last an additional 2 h beforethe images were ready for viewing.

Despite these difficulties and the problems associatedwith servicing the prototype, the clinical evaluationprogramme established some important criteria for hard-ware fusion imaging. These results somewhat optimisti-cally suggested a 30% change in oncology patientmanagement using PET/CT images compared with PETalone [4, 5]. The routine availability of fused anatomy andfunction for every patient imaged on the prototype oftenconvinced referring physicians to request PET/CT directlyrather than PET alone. Enthusiastic referrals rapidlyexceeded the capacity of the prototype that had initiallybeen envisaged as a device to establish proof-of-principle,not one intended for routine clinical operation.

Conflicting attitudes

As the images from the PET/CT prototype spread throughpresentation at national and international meetings andthrough publications, the demand for a commercial PET/CT design grew. At the Society of Nuclear Medicineannual meeting in Los Angeles in 1999, Dr. Henry Wagnerselected a head and neck cancer study from the PET/CTprototype as the Image of the Year during his annualHighlights Lecture. This recognition helped raise aware-ness of the potential role of combined PET/CT imaging at atime when the prototype was the only such device inexistence. Awareness was further increased when the PET/CT was selected as the Medical Invention of the Year byTIME Magazine in December 2000 [6]. While such initialrecognition was important, the real impetus for PET/CTcame from the routine availability of co-registered anatomyand function and the gradual recognition that PET/CT isconsiderably more than just CTand PET. It is, in fact, a newimaging modality.

In response to growing demand from the medicalcommunity, the first commercial PET/CT design wasshowcased at the Society of Nuclear Medicine meeting inSt. Louis in June 2000, with initial installations appearingin major medical centres by May 2001. Five manufacturersnow offer close to 20 different PET/CT designs and withinless than 5 years the number of installed units has risen tomore than 1,000. The adoption of PET/CT as a replacementfor PET-only, particularly in oncology, has been impres-sive, with a growth rate rarely seen before in the medicalimaging field. The impact has been profound, far

exceeding that of simply providing co-registered anatom-ical and functional images for all patients.

Inevitably, combined PET/CT imaging has not beenwithout its critics. It has variously been described asdisruptive technology [7] and accused of heralding thedeath of nuclear medicine. Early presentations of theresults from the prototype were met with a mixture ofenthusiasm and concern—enthusiasm for the quality of thefused images and the availability of precise anatomicalinformation with every PET scan; concern because of thecomplications related to reading and interpretation of fusedimages, and ownership and operation of a dual-modalityscanner. The situation can be summarised by the commentthat fusing technologies is one thing, fusing medicalspecialties is another, the implication being that, though theformer may have been achieved, the latter would be evenmore challenging. A few rather defensive articles havebeen published suggesting that PET/CT imaging wouldlikely perform no better than retrospective image fusion[8, 9].

Despite some of these initial reservations, the adoptionof PET/CT has been surprisingly rapid from the timecommercial designs became available. Within 2–3 years ofthe introduction of PET/CT, few if any PET scanners werebeing sold by the major vendors, although there wereessentially no published studies that established a supe-riority of PET/CT over PET for any particular cancer. Eventoday, 5 years after the introduction of PET/CT into theclinical arena, there are few evidence-based studiescomparing PET with PET/CT, or comparing PET and CTsoftware fusion with PET/CT. The reasons for this aremainly methodological: it would be difficult to justifyexposure of a patient to radiation for multiple scans in orderto make such a comparison prospectively. In any case, theavailability of an intrinsically co-registered CT will, atworst, have no effect on the established sensitivity of thePET study, and at best will improve both sensitivity andspecificity.

Software fusion certainly does not fill the role forwhole-body imaging, where the most sophisticated regis-tration algorithms have difficulty aligning the CT and PETimage sets [10–12]. Even when local registration isachieved by software, accurate alignment of the entirebody is complicated by different patient positioning,different scanner bed profiles, different disease statesbecause of the temporal separation between the scans, and,most importantly, the uncontrolled movement of internalorgans that occurs between separate scans. Above all, thereis no accepted gold standard for retrospective softwareregistration. Combined PET/CT has thus introduced a newstandard for fusion accuracy against which softwaretechniques can now be compared.

It has been suggested by some that the adoption of PET/CT is driven by marketing and vendor pressure and is notbased on any sound scientific evidence or medicalprinciples. While, as stated, it is true that as yet there areno extensive prospective studies establishing the supe-riority of PET/CT over either PET alone or software fusion,

858

European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 8, August 2006

Page 3: Putting ‘clear’ into nuclear medicine: a decade of PET/CT development

UNCO

RREC

TEDPR

OOF

the availability of co-registered detailed anatomy whenreading the PET study simply makes good clinical practice.Increasingly, there is documentary evidence that interpret-ing PET images with the associated anatomical scan notonly improves the accuracy of the PET reading, but alsoassists the radiologist to focus on regions of abnormal FDGuptake; the confidence of the reader, whether radiologist ornuclear physician, is greatly increased by having access toco-registered images [13, 14]. Thus, PET/CT imagingcombines the power of a functional image with that of ananatomical image, such that the whole is greater than thesum of the parts.

Why, then, has the introduction of PET/CT created suchcontroversy? Much of it is related to artificial territorial andownership issues that will eventually be resolved to thebenefit of the patient. Other concerns that have fuelled thecontroversy include cost and reliability of the technology,and whether there is a real clinical need for combined PET/CT imaging. On the positive side, PET/CT has broughtfunctional imaging to the forefront in radiology, awakeningthe radiological community to the power of imagingmetabolic processes within an anatomical context. (Indeed,many radiologists would consider a return to PET-onlyimaging in oncology unthinkable.) However, assessment offunction is much more than just the mapping of glucoseutilisation through the uptake of FDG. The strength ofthe tracer technique lies in the range of potentialbiomarkers for imaging human physiology, includingexisting biomarkers and those that are still under activedevelopment. As the biomarkers become even morespecific for a particular metabolic pathway, co-registrationof functional and anatomical information will becomecritical, as will the close collaboration between radiologyand nuclear medicine.

A survey of recent PET/CT publications has suggestedan incremental clinical benefit of PET/CT over PET alonein 10% of patients scanned. The obvious question iswhether a 10% change in clinical patient management, thatis based on FDG as the biomarker of choice, is enough tojustify the adoption of a new imaging modality, keeping inmind that for other biomarkers the change in patientmanagement could be even more significant. Obviously, ifthe 10% group could be identified proactively, a differentimaging strategy could be applied, possibly without theneed for PET/CT. Apart from making good clinical sense,PET/CT offers a number of other important advantages,including significantly reduced scan times and higherpatient throughput [15] with the use of CT-based attenu-ation correction [16], and increased patient and physicianconvenience due to the requirement for only one scan forboth modalities. In addition, a single integrated report fordiagnostic or therapeutic purposes can be provided to thereferring physician whenever the CT is performed accord-ing to a clinical protocol.

The evolution of PET/CT

Since the introduction of PET/CT in 2001, the technologyhas progressed in line with advances in CT and PETseparately [17]. Assuming that PET/CT will likely becomethe imaging modality of choice in oncology, how shouldthe technology improve to meet the future demands ofmedical oncologists, surgeons, radiation therapists andothers? Some possible future trends, including operationaland methodological improvements, may be summarised asfollows:

PET

In combined PET/CT imaging, the PETcomponents are thelimiting factors in terms of spatial, temporal and contrastresolution. Recent progress in detector design has led tofaster scintillators and smaller detector elements, offeringimproved count rate performance and better spatial reso-lution [18, 19]. Faster coincidence electronics coupled withthese new scintillators yield reduced randoms and scatterrates and thus an improved signal-to-noise ratio. The newscintillators also offer the possibility of time-of-flightacquisition, where the time difference between the arrivalof the two annihilation photons is used to localise theannihilation point; the better the time resolution, the moreaccurately the annihilation point can be localised, whichthen leads to a further increase in the signal-to-noise ratio[20]. A significant improvement in overall system sensi-tivity can be attained by further extending the axial field ofview coverage, particularly as current PET systems collectless than 2% of the annihilations occurring within the fieldof view owing to the small solid angle coverage. Three-dimensional acquisition [21] that increases the solid anglecoverage has been a hotly debated issue for whole-bodyimaging over the past decade [22–25]. The new scintilla-tors and electronics address many of the concerns related to3D acquisition and therefore 3D will likely becomestandard in the future, particularly in view of thedesirability of reducing the scan duration by makingmaximum use of the emitted radiation.

CT

After several years of relatively slow progress, the recentadvances in CT technology have been dramatic [26]. Fromsingle-slice, state-of-the-art CT scanners in 2000, recentdesigns incorporate 64 slices with the promise of 128 oreven 256 to come. New X-ray tube technology and CTscanner designs with dual tubes result in shorter imagingtimes and novel acquisition schemes [27]. However, suchhigh CT performance, driven primarily by cardiac appli-cations, exceeds that required for oncology imaging, the

859

European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 8, August 2006

Page 4: Putting ‘clear’ into nuclear medicine: a decade of PET/CT development

UNCO

RREC

TEDPR

OOF

most widespread usage for PET/CT, and therefore thedesire for disease-specific applications may justify modulardesigns.

Combined PET/CT

A major benefit of PET/CT has been the use of CT imagesfor attenuation correction of PET data. Simple bi-linearscaling models that originated from early work on dual-modality imaging have now been validated to transform theimages acquired at the X-ray energy (∼70 keV) up to thePET annihilation photon energy of 511 keV [28–30].Extensive clinical studies have demonstrated the accuracyof CT-based attenuation and the methodology has nowcompletely replaced the use of the PET transmissionsources that were required in the previous PET tomo-graphs. The advantages of CT-based attenuation—lownoise, high contrast and rapid scanning protocols—willensure that it remains the method of choice for attenuationcorrection in all PET/CT studies, assuming that therequired correction procedures (e.g. [31]) are providedwith the clinical systems.

Dose will continue to be a concern for PET/CTscanning, and particularly the dose from the CT examina-tion. Strategies for reducing radiation dose from CT arewell known and include tube current modulation, special-ised filtering procedures and education to select applica-tion-specific acquisition parameters. Many of today’s PET/CT procedures are simply based on the same PET and CTscan protocols that are used with the separate systems.Combined protocols must be developed for specificapplications, particularly where the CT protocol involvesthe use of intravenous or oral contrast media [32].

Of even greater importance is to ensure that themodality is used properly and for the benefit of the patient[33]. Since, for most people, 80% of their lifetime exposureto ionising radiation will originate from diagnostic medicalprocedures, it is time to move beyond subjecting the patientto all imaging procedures and then deciding which ones aredemanded retrospectively. A more logical approach isneeded that is driven by patient care issues and not byfinancial or legal considerations. For example, referringphysicians must be educated to know when to request fullclinical PETand CTscans and when a PETscan with a low-dose CT scan will be adequate [34, 35].

It is already evident that the strength of PET/CT will bein treatment planning and the assessment of response at theearliest possible stage to give the patient the best possibleprognosis. Since assessment of treatment response requirescomparison of PET/CT studies acquired at different timesduring the course of treatment, co-registering these datasets could still be an important application for softwarefusion algorithms [12, 36]. The routine availability ofintrinsically aligned anatomical and functional imagesoffers the possibility of new metrics for definition oftreatment response that go beyond the conventionalindependent estimates of size and standardised uptakevalues [37].

Concluding remarks

By bringing together morphology and molecular imaging,PET/CT—an evolution in imaging technology—has cre-ated something of a revolution in diagnostic imaging.Morphology without function is nothing more than alifeless corpse, whereas function without structure is asephemeral as a ghost. By adding morphology to molecularimaging, form to function, the whole is unquestionablygreater than the sum of the parts.

As with any new imaging modality, the key to itseffective use is education: educating the technologists tooperate a dual-modality tomograph efficiently; educatingthe referring physicians to request the appropriate studies;educating the interpreting physicians to make the bestpossible use of the available information from the scan;educating hospital administrators on the benefits of PET/CT to the institution; educating vendors to provide efficientinterface and adequate viewing platforms; and educatinghealthcare professionals on the cost-effectiveness ofclinically indicated multi-modality imaging. These arenot straightforward issues to resolve but they urgently needto be addressed. Great progress has been made in the 5years since the introduction of the first commercial PET/CT system, but much remains to be done. Inevitably, thereal problems are often obscured by territorial attitudes anddefensive reactions that are aimed more at maintaining theprofessional status quo than at integrating a new modalityinto clinical care for the benefit of the patient. However, asa patient one desires the best and most accurate diagnosisas quickly as possible. There, PET/CT can help. However,to be effective, a cooperative approach is required by thedisciplines involved: referring physicians, radiologists,nuclear medicine physicians, physicists, technologists andhospital administrators. In contrast to the opinion quotedearlier, PET/CT will not be the death of nuclear medicinebut a catalyst that brings it to the forefront of medicaldiagnostic practice. PET/CT will overcome territorial andprotective practices; that, PET/CT will most certainly do.

References

1. Townsend DW, Cherry SR. Combining anatomy with function:the path to true image fusion. Eur Radiol 2001;11:1968–74

2. Townsend D, Beyer T, Kinahan P, et al. The SMART scanner: acombined PET/CT tomograph for clinical oncology. Radiology1998;209(P):169–70

3. Beyer T, Townsend DW, Brun T, Kinahan PE, Charron M,Roddy R, et al. A combined PET/CT tomograph for clinicaloncology. J Nucl Med 2000;41:1369–79

4. Charron M, Beyer T, Bohnen NN, Kinahan PE, Dachille M,Jerin J, et al. Image analysis in patients with cancer studiedwith a combined PET and CT scanner. Clin Nucl Med 2000;25:905–91

5. Kluetz PG, Meltzer CC, Villemagne VL, Kinahan PE, ChanderS, Martinelli MA, et al. Combined PET/CT imaging inoncology: impact on patient management. Clinical PositronImaging 2000;3:223–30

6. Jaroff L. A winning combination. TIME Magazine 2000;156(23):72–4

860

European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 8, August 2006

Page 5: Putting ‘clear’ into nuclear medicine: a decade of PET/CT development

UNCO

RREC

TEDPR

OOF

7. Wiley G. Disruptive technology: the conflict over PET/CT.Decis Imaging Econ 2005;June Issue:17–20

8. Jager P, Slart R, Corstens F, Oyen W, Hockstra O, Teule J. PET-CT: a matter of opinion. Eur J Nucl Med Mol Imaging2003;30:470–1

9. Lemke A-J, Niehues SM, Hosten N, Amthauer H, Boehmig M,Stroszczynski C, et al. Retrospective digital image fusion ofmultidetector CT and 18F-FDG PET: clinical value in pancreaticlesions—a prospective study with 104 patients. J Nucl Med2004;45:1279–86

10. Slomka PJ. Software approach to merging molecular withanatomic information. J Nucl Med 2004;45:36S–45S

11. Shekhar R, Walimbe V, Raja S, Zagrodsky V, Kanvinde M, WuG, et al. Automated 3-dimensional elastic registration of whole-body PET and CT from separate or combined scanners. J NuclMed 2005;46:1488–96

12. Pietrzyk U. Does PET/CT render software fusion obsolete?Nuklearmedizin 2005;44:S13–7

13. Pannu HK, Bristow RE, Cohade C, Fishman EK, Wahl RL.PET-CT in recurrent ovarian cancer: initial observations.Radiographics 2004;24:209–23

14. Branstetter BF 4th, Blodgett TM, Zimmer LA, Snyderman CH,Johnson JT, Raman S, et al. Head and neck malignancy: is PET/CT more accurate than PET or CT alone? Radiology2005;235:580–6

15. von Schulthess GK. Cost considerations regarding an integratedCT-PET system. Eur Radiol 2000;10:S377–80

16. Kinahan PE, Townsend DW, Beyer T, Sashin D. Attenuationcorrection for a combined 3D PET/CT scanner. Med Phys1998;25:2046–53

17. Townsend D, Carney JP, Yap JT, Hall NC. PET/CT today andtomorrow. J Nucl Med 2004;45:4S–14S

18. Townsend D, Beyer T. Anato-molecular imaging: combiningstructure and function. In: Bailey DL, Townsend DW, Valk PE,Maisey MN, editors. Positron emission tomography: principlesand practice. London: Springer; 2005. p 179–202

19. Pfannenberg A, Eschmann S, Brechtel K, et al. A new high-resolution, multi-slice PET/CT tomograph for state-of-the-artoncology imaging—radiologist’s perspective. J Nucl Med2005;46:414P

20. Moses W. Time of flight in PET revisited. IEEE Trans Nucl Sci2003;50:1325–30

21. Townsend DW, Isoardi RA, Bendriem B. Volume imagingtomographs. In: Townsend DW, Bendriem B, editors. Theoryand practice of 3D PET. Dordrecht: Kluwer Academic; 1998.p 111–32

22. Lodge MA, Badawi RD, Gilbert R, Dibos PE, Line BR.Comparison of 2-dimensional and 3-dimensional acquisitionfor 18F-FDG PET oncology studies performed on an LSO-based scanner. J Nucl Med 2006;47:23–31

23. Visvikis D, Griffiths D, Costa D, Bomanji J, Ell P. Clinicalevaluation of 2D versus 3D whole-body PET image qualityusing a dedicated BGO PET scanner. Eur J Nucl Med MolImaging 2005;32:1050–6

24. Votaw JR, White M. Comparison of 2-dimensional and 3-dimensional cardiac 82Rb PET studies. J Nucl Med 2001;42:701–6

25. Ferreira NC, Trébossen R, Bendriem B. Assessment of 3-DPET quantitation: influence of out of the field of viewradioactive sources and of attenuating media. IEEE TransNucl Sci 1998;45:1670–5

26. Kalender WA. CT: the unexpected evolution of an imagingmodality. Eur Radiol 2005;15:D21–4

27. Flohr T, McCollough C, Bruder H, Petersilka M, Gruber K,Suss C, et al. First performance evaluation of a dual-source CT(DSCT) system. Eur Radiol 2006;16:256–68

28. Burger C, Goerres G, Schoenes S, Buck A, Lonn AHR,Schulthess GKv. PET attenuation coefficients from CT images:experimental evaluation of the transformation of CT into PET511-keV attenuation coefficients. Eur J Nucl Med Mol Imaging2002;29:922–92

29. Kinahan P, Hasegawa B, Beyer T. X-ray based attenuationcorrection for PET/CT scanners. Semin Nucl Med 2003;XXXIII:166–79

30. Watson CC, Rappoport V, Faul D, Townsend DW, Carney JPJ.A method for calibrating the CT-based attenuation correction ofPET in human tissue. IEEE Trans Nucl Sci 2006; in press

31. Beyer T, Bockisch A, Kuhl H, Martinez M-J. Whole-body18F-FDG PET/CT in the presence of truncation artifacts. J NuclMed 2006;47:91–9

32. Brechtel K, Klein M, Vogel M, Mueller M, Aschoff P, Beyer T,et al. Optimized contrast enhanced CT protocols for diagnosticwhole-body 18F-FDG PET/CT: single-phase versus multi-phaseCT imaging. J Nucl Med 2006;47(3):470–6

33. Bockisch A, Beyer T, Antoch G, Freudenberg LS, Kuhl H,Debatin JF, et al. Positron emission tomography/computedtomography—imaging protocols, artifacts, and pitfalls. MolImaging Biol 2004;6:188–99

34. Kuehl H, Antoch G. How much CT do we need for PET/CT? Aradiologist’s perspective. Nuklearmedizin 2005;44:S24–31

35. Strobel K, Thuerl CM, Hany T. How much intravenous contrastis needed in FDG-PET/CT? Nuklearmedizin 2005;44:S32–7

36. Beyer T. Towards truly integrated hardware fusion with PET/CT. Nuklearmedizin 2005;44:S5–S12

37. Therasse P, Arbuck SG, Eisenhauer EA, Wanders J, KaplanRS, Rubinstein L, et al. New guidelines to evaluate the re-sponse to treatment in solid tumors. J Natl Cancer Inst 2000;92:205–16

861

European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 8, August 2006