This Article Abstract Figures Only Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Townsend, D W Articles by Beyer, T Search for Related Content PubMed PubMed Citation Articles by Townsend, D W Articles by Beyer, T

A combined PET/CT scanner: the path to true image fusion

¹ University of Pittsburgh, Pittsburgh, Pennsylvania, USA and ² CPS Innovations, Knoxville, Tennessee, USA

Correspondence: DW Townsend, PET Facility, Department of Radiology, University of Pittsburgh, 200 Lothrop St, Pittsburgh, PA 15213, USA. email: townsenddw@msx.upmc.edu.

	Abstract

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

 
Software-based image fusion is used routinely for the alignment of functional and anatomical images of the brain. For other parts of the body, image registration is more problematic owing to differences in patient positioning, scanner bed profiles and the involuntary movement of internal organs. An alternative to the software approach is a scanner that acquires both function and anatomy during a single imaging session: a fusion of the technologies rather than a fusion of the images post hoc. Consequently, we designed and built a prototype combined PET and CT scanner comprising a clinical CT and a clinical PET scanner mounted together in a single gantry. Over 300 cancer patients have been imaged in the scanner to establish the clinical value of the combined PET/CT approach. The CT images were used to provide essentially noiseless attenuation correction factors for the PET data. The widespread interest created by the patient studies acquired with the prototype PET/CT stimulated commercial activity and several major vendors of medical imaging equipment now offer combined PET/CT designs. This paper reviews the development of the combined PET/CT scanner, and illustrates the clinical aspects with some typical studies in cancer patients. The potential impact on medical practice of the commercial availability of PET/CT scanner technology at affordable cost is assessed.

	Introduction

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

 
This paper briefly reviews the development of a combined PET/CT scanner designed primarily for applications in clinical oncology. The PET/CT design work began with a grant from the National Cancer Institute in 1995 and the prototype was completed in April 1998 and installed at the University of Pittsburgh Medical Center (UPMC). The prototype was operated up to August 2001 when it was replaced with a commercial PET/CT scanner, the biograph (Siemens Medical Solutions). The recent extended reimbursement by major insurance carriers in the US of PET studies for a range of different cancers has led to a rapid growth in the clinical demand for PET. Widespread recognition of the importance of imaging anatomy and function together, based to a large extent on the studies from the prototype PET/CT, created a demand for combined PET/CT scanners for imaging cancer, a demand that is now being met by a number of manufacturers offering commercial PET/CT designs. It may, however, be a while before the advantages of the combined PET/CT technology are fully established.

	Fusion imaging: anatomy and function

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

 
The advantages of fused anatomical and functional images have long been appreciated, where the fusion is achieved by software methods. While generally successful for the brain [1], software approaches often encounter significant difficulties with the rest of the body. Routine fusion software requires access to images from different modalities, and the alignment procedures are generally labour intensive and uncertain of success. Alignment algorithms specifically use a mutual information content measure of the two image sets and will generally fail to converge when the content of the two sets is highly dissimilar. This will arise typically when either the functional image contains little or no correlative anatomical information or when the anatomical information in the functional image represents a different anatomy owing to, for example, patient movement or discrepancies in patient positioning between the two scans. Problems due to patient movement include involuntary and uncontrollable movement of internal organs that inevitably arise when a patient is imaged on different scanners and at different times, even when care is taken to ensure the same external position of the patient.

While some of the patient movement and positioning errors may be overcome by alignment using non-linear image warping techniques and 3D elastic transformations, an alternative approach to post hoc image fusion is actually to fuse, or combine, the imaging technologies into one scanner, a scanner that can acquire accurately aligned anatomical and functional images in the same scanning session [2]. The advantages of fusing the technologies are many, while the challenges are considerable [3]. The two imaging technologies selected for this development were PET for the functional component and CT for the anatomical component. PET is now widely used to image cancer owing to the uptake of ¹⁸F-deoxyglucose (FDG), a PET tracer, by malignant cells. CT is still the anatomical imaging modality of choice in over 85% of cancer patients. It is also used extensively for therapy monitoring by assessing changes in lymph node size, and for radiation therapy treatment planning. As an additional advantage, when CT is combined with PET, the CT images can be scaled in energy and used to correct the PET data for attenuation effects [4], thus eliminating the requirement for a separate time-consuming PET transmission scan.

	PET/CT prototype design

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

 
The PET/CT prototype [2] is based on combining a spiral CT scanner (Somatom AR.SP) with the PET components from a rotating partial ring tomograph, the ECAT ART [5]. The ECAT ART is a partial ring, rotating tomograph that comprises dual arrays of BGO block detectors. The detector blocks are 54 mm x 54 mm x 20 mm in size, cut into 8 x 8 crystals each of dimension 6.75 mm x 6.75 mm x 20 mm. The axial FOV is therefore 16.2 cm (24 partial rings of 6.75 mm). Additional shielding of the PET components from out-of-field activity is provided by arcs of lead, 2.5 cm thick, mounted on both sides of the detector assembly and projecting 8.5 cm into the FOV beyond the front face of the detectors.

The CT is a third generation helical scanner, a Somatom AR.SP. Spiral, or helical, CT acquires multiple axial slices by a continuous motion of the patient bed. This results in shorter scan times and lower overall dose to the patient. The CT scanner has a metal ring M-CT 141 tube that produces X-ray spectra of 110 kV_p and 130 kV_p with a 6.5 mm Al-equivalent filter. The tube is operated with a flying spot, and thus 1024 detectors can be read from 512 xenon gas-filled Quantillarc chambers. The X-ray tube, cooling system, detectors and readout electronics are all mounted on the rotating support of the CT scanner.

The packing density of the CT components precludes the possibility to mount the PET detectors on the same side of the rotating support as the CT. Instead, the PET components are mounted to the rear of the CT support ring, on a separate aluminum annulus attached to the CT support. The PET components include the detectors and electronics, coincidence processor and the optically coupled data transmitters. An asynchronous motor provides rotation of the entire assembly of PET and CT components at 30 rpm. The device is housed inside a single gantry 170 cm wide, 168 cm high and 110 cm deep. The centres of the two tomographs are axially offset by 60 cm. A common patient bed is installed at the front of the combined gantry. Dual-modality PET and CT images can be acquired for a total axial extent of 100 cm, sufficient to cover the range for most patients from chin to lower thigh. A schematic diagram of the prototype design is shown in Figure 1.

View larger version (53K): [in this window] [in a new window]   Figure 1. Schematic diagram of the combined PET/CT prototype design that was based on a Siemens Somatom AR.SP CT scanner and an ECAT ART PET scanner.

	PET/CT designs: from prototype to commercial scanner

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

View larger version (50K): [in this window] [in a new window]   Figure 2. The biograph commercial combined PET/CT scanner (Siemens Medical Solutions).

	Clinical imaging with the biograph

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

 
During its 3 year life span, the prototype PET/CT at the University of Pittsburgh established the importance of combined functional and anatomical imaging for diagnosis, staging and therapy monitoring in oncology. Details of the clinical studies acquired on the prototype can be found elsewhere [6–8]. In August 2001, the prototype was replaced with the commercial biograph PET/CT (Siemens Medical Solutions) (Figure 2). The biograph in Pittsburgh comprises a single-slice Siemens Emotion CT scanner in tandem with an ECAT HR+ PET scanner. The HR+ has no septa and operates entirely in 3D mode. The transmission rod sources were also removed and attenuation correction is entirely CT-based. A new patient bed design allows a combined scan range for both PET and CT of 145 cm. The patient pallet is attached at one end to a support pedestal that moves on floor-mounted rails. Since the cantilever point is fixed, the problem of increasing bed deflection with increasing distance into the scanner is eliminated. A flat pallet is available for radiation therapy patients, and the entry port diameter is 70 cm for both PET and CT, greatly facilitating the positioning of such patients. While the overall tunnel length is 110 cm, the open, contoured design and large patient port essentially eliminate problems with claustrophobia. Unlike the prototype, both PET and CT acquisition and reconstruction run under a single protocol on one workstation.

To date, over 600 patients have been scanned on the biograph at UPMC with excellent results. As shown in Figure 3, a typical PET/CT acquisition protocol begins with a 370 MBq injection of FDG and a 60 min uptake period. The patient is then positioned in the scanner with the first transaxial section to be imaged aligned with the FOV of the CT. An initial scout scan (survey) is performed to determine the axial range of the spiral scan. For the single-slice Somatom Emotion scanner, a maximum 100 cm spiral scan can be acquired in 90 s in a single acquisition. Patients are allowed to breathe shallowly during the CT scan to match the breathing cycle during the PET scan that cannot be acquired with breath-hold. While such a protocol may introduce breathing artefacts into the CT image, it is more comfortable for the patient and helps to reduce the mismatch between the CT and PET images. Once the spiral CT scan is completed, the patient bed is automatically moved to the start position of the multi-bed PET acquisition, and the PET scan is initiated. For a fixed total scan time, an emission time of 4–7 min per bed position is selected depending on the specific clinical protocol and the patient size. The ranges scanned for PET and CT are automatically matched to ensure that CT-based attenuation correction factors are available for all PET sections, and to avoid unnecessary X-ray exposure of the patient in regions where PET emission data are not acquired. As shown in Figure 3, the CT images are used for attenuation and scatter correction, and the corrected PET data are reconstructed with Fourier rebinning (FORE) and the attenuation-weighted ordered subset EM algorithm (AWOSEM). Complete acquisition of both PET and CT data takes less than 30 min and the fused images are available for viewing within 5 min of the completion of the scan. The images are viewed on a separate fused image display station.

View larger version (39K): [in this window] [in a new window]   Figure 3. A typical combined PET/CT protocol for a patient study acquired on the biograph. For the CT scan, the pitch is 1.1.

View larger version (39K): [in this window] [in a new window]   Figure 4. A frontal section from a PET/CT scan of a 54-year-old female with a history of cervical cancer; the coronal sections show (a) the CT, (b) the FDG-PET and (c) the fused PET/CT image.

View larger version (75K): [in this window] [in a new window]   Figure 5. A 74-year-old male with a history of colorectal cancer: (a) a coronal PET image showing focal FDG uptake in a right anterior para-aortic lymph node, (b) the transaxial CT image (upper) and fused PET/CT image (lower) taken at the level indicated by the line in (a), and (c) the fused PET/CT image corresponding to the PET image in (a). The patient was also found to have a lesion in the liver as shown in (d) the coronal PET image, (e) the CT image (upper) and fused PET/CT image (lower) taken at the level indicated by the line in (d), and (f) the fused PET/CT image corresponding to the PET image in (d).

Case 2
A 73-year-old male with a history of colorectal cancer had undergone a sigmoidectomy and chemotherapy. A clinical PET scan 2 months prior to the PET/CT had been read as normal. A recent CT scan suggested a possible liver lesion and an abnormal para-aortic node. The PET/CT scan clearly showed intense uptake of FDG in a right anterior para-aortic lymph node (Figure 5a,b,c), and in a liver lesion (Figure 5d,e,f), both consistent with malignancy.

	Future developments

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

 
The current PET/CT designs from the major manufacturers comprise a commercial CT scanner in tandem with a commercial PET scanner. The level of physical integration is actually less than that of the original prototype where the CT and PET components were mounted on the same rotating support [2]. Based on the positive clinical experience with PET/CT at UPMC and other institutions, there is going to be a demand for a reduction in cost of these devices and for a greater level of integration. This may obviously be achieved through the design of a scanner specifically for combined anatomical and functional imaging, rather than combining separate CT and PET scanners, as in the current approaches. By avoiding the duplication of data acquisition and image reconstruction functions, for example, a more integrated design should also allow cost savings over current commercial PET/CT scanners. The goal is then to design and build a device specifically for imaging the function and anatomy of cancer in the most informative and effective way, without conceptualizing it as combined PET and CT. The development of devices specifically for imaging a particular disease (e.g. cancer) differs from the conventional approach of, for example, an all-purpose anatomical imaging device such as a CT scanner. This new concept relates more to a disease management approach than the usual subdivision into medical specialties such as radiology and nuclear medicine.

For such an approach to succeed, the new designs must be cost-effective and reliable [9]. The rapidly increasing use of functional imaging such as PET in areas that have traditionally been dominated by anatomical imaging modalities will demand reliable and easy-to-use PET/CT scanners that can achieve high throughput. The recent introduction of fast scintillators such as LSO and GSO for PET detectors is occurring at just the right moment for PET/CT where a reduction in the lengthy PET imaging time is required, more closely matching that of the CT. While it is unlikely that the PET imaging time will be reduced to the 30–60 s or so required for CT scanning, an order of magnitude reduction is to be expected with new high-performance LSO detectors. For example, it is anticipated that an LSO scanner with integrated CT could achieve an overall whole-body scan time of less than 10 min. Such a scanner would represent a breakthrough in cancer imaging, eliminating problems of patient movement and substantially reducing artefacts due to respiration. Throughput would increase, as would patient comfort and convenience. New applications, such as dynamic whole-body scans and the use of short-lived radioisotopes (e.g. ¹¹C) would then be within reach.

Future developments in combined PET/CT scanners will be exciting, attaining a higher level of integration and anatomical and functional imaging performance than ever before. By playing an important role, not only in diagnosis and staging of cancer, but in designing and monitoring appropriate therapies, the combined PET/CT scanner will have a significant impact on patient care, survival and quality of life.

	Acknowledgments

 
The combined PET/CT project has involved many people at the University of Pittsburgh and CPS Innovations over the last few years. In particular, we acknowledge the seminal contribution of Dr Ron Nutt who led the PET/CT prototype development team at CPS Innovations and Dr Charles Watson who led the development of the CPS PET/CT scanner. The PET instrumentation and methodology group at the University of Pittsburgh made a major contribution to this project, in particular Drs Paul Kinahan, Jonathan Carney, David Brasse and Jeff Yap. Finally we thank the many physicians and technologists in the Pittsburgh PET Facility who contributed to the PET/CT clinical evaluation program, especially Drs Charron, Meltzer, Blodgett and McCook, and technologists Denise Ratica, Stacey Mckenzie, Marsha Martinelli and Donna Mason. The PET/CT development project is supported by a National Cancer Institute grant, CA 65856.

	References

Top Abstract Introduction Fusion imaging: anatomy and... PET/CT prototype design PET/CT designs: from prototype... Clinical imaging with the... Future developments References

 

1. Woods RP, Mazziotta JC, Cherry SR. MRI-PET registration with an automated algorithm. J Comp Assist Tomogr 1993;17:536–46.[Medline]
2. Beyer T, Townsend DW, Brun T, Kinahan PE, Charron M, Roddy R, Jerin J, Young J, Byars L, Nutt R. A combined PET/CT scanner for clinical oncology. J Nucl Med 2000;41:1369–79.[Abstract/Free Full Text]
3. Townsend DW. A combined PET/CT scanner: the choices. J Nucl Med 2001;42:533–4.[Free Full Text]
4. Kinahan PE, Townsend DW, Beyer T, Sashin D. Attenuation correction for a combined 3D PET/CT scanner. Med Phys 1998;25:2046–53.[Medline]
5. Townsend DW, Beyer T, Jerin J, Watson CC, Young J, Nutt R. The ECAT ART scanner for positron tomography: 1. Improvements in performance characteristics. Clin Posit Imaging 1999;2:5–15.
6. Kluetz PG, Meltzer CC, Villemagne MD, Kinahan PE, Chander S, Martinelli MA, Townsend DW. Combined PET/CT imaging in oncology: impact on patient management. Clin Posit Imaging 2001;3:31–8.
7. Townsend DW, Beyer T, Kinahan PE, et al. Recent studies with a combined PET/CT scanner: a synergistic approach to patient management. In: Tamaki N, Tsukamoto E, editors. Positron emission tomography in the Millenium. Amsterdam: Elsevier, 2000:229–44.
8. Charron M, Beyer T, Bohnen N, Kinahan PE, Dachille M, Jerin J, Nutt R, Meltzer CC, Villemagne V, Townsend DW. Image analysis in patients with cancer studied with a combined PET and CT scanner. Clin Nucl Med 2000;25:905–10.[Medline]
9. Townsend DW, Cherry SR. Combining anatomy with function: the path to true image fusion. Eur Radiol 2001;11:1968–74.[Medline]

This article has been cited by other articles:

				H.-Y. Lee, Z. Li, K. Chen, A. R. Hsu, C. Xu, J. Xie, S. Sun, and X. Chen PET/MRI Dual-Modality Tumor Imaging Using Arginine-Glycine-Aspartic (RGD)-Conjugated Radiolabeled Iron Oxide Nanoparticles J. Nucl. Med., August 1, 2008; 49(8): 1371 - 1379. [Abstract] [Full Text] [PDF]

				B. J. Pichler, H. F. Wehrl, and M. S. Judenhofer Latest Advances in Molecular Imaging Instrumentation J. Nucl. Med., June 1, 2008; 49(Suppl_2): 5S - 23S. [Abstract] [Full Text] [PDF]

				W. Cai and X. Chen Multimodality Molecular Imaging of Tumor Angiogenesis J. Nucl. Med., June 1, 2008; 49(Suppl_2): 113S - 128S. [Abstract] [Full Text] [PDF]

				B. Rodriguez-Vigil, N. Gomez-Leon, I. Pinilla, D. Hernandez-Maraver, J. Coya, L. Martin-Curto, and R. Madero PET/CT in Lymphoma: Prospective Study of Enhanced Full-Dose PET/CT Versus Unenhanced Low-Dose PET/CT J. Nucl. Med., October 1, 2006; 47(10): 1643 - 1648. [Abstract] [Full Text] [PDF]

				B. J. Pichler, M. S. Judenhofer, C. Catana, J. H. Walton, M. Kneilling, R. E. Nutt, S. B. Siegel, C. D. Claussen, and S. R. Cherry Performance Test of an LSO-APD Detector in a 7-T MRI Scanner for Simultaneous PET/MRI J. Nucl. Med., April 1, 2006; 47(4): 639 - 647. [Abstract] [Full Text] [PDF]

				M. Farsad, R. Schiavina, P. Castellucci, C. Nanni, B. Corti, G. Martorana, R. Canini, W. Grigioni, S. Boschi, M. Marengo, et al. Detection and Localization of Prostate Cancer: Correlation of 11C-Choline PET/CT with Histopathologic Step-Section Analysis J. Nucl. Med., October 1, 2005; 46(10): 1642 - 1649. [Abstract] [Full Text] [PDF]

				T. Rose Educating the Hybrid Technologist Radiol. Technol., September 1, 2005; 77(1): 15 - 18. [Full Text] [PDF]

				L. Kostakoglu, R. Hardoff, R. Mirtcheva, and S. J. Goldsmith PET-CT Fusion Imaging in Differentiating Physiologic from Pathologic FDG Uptake RadioGraphics, September 1, 2004; 24(5): 1411 - 1431. [Abstract] [Full Text] [PDF]

				T. Hamaoka, J. E. Madewell, D. A. Podoloff, G. N. Hortobagyi, and N. T. Ueno Bone Imaging in Metastatic Breast Cancer J. Clin. Oncol., July 15, 2004; 22(14): 2942 - 2953. [Abstract] [Full Text] [PDF]

				J B Bingham Where can FDG-PET contribute most to anatomical imaging problems? Br. J. Radiol., November 1, 2002; 75(90009): S39 - 52. [Full Text] [PDF]

				P K Marsden, D Strul, S F Keevil, S C R Williams, and D Cash Simultaneous PET and NMR Br. J. Radiol., November 1, 2002; 75(90009): S53 - 59. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Townsend, D W Articles by Beyer, T Search for Related Content PubMed PubMed Citation Articles by Townsend, D W Articles by Beyer, T