Principles and Practice of PET/CT

EANMEANM® Eu ne rope edici an Association of Nuclear M

Publications · Brochures

Part 1 A Technologist‘s Guide

Produced with the kind Support of Editors

Peter Hogg Giorgio Testanera Head of Diagnostic Imaging Research Istituto Clinico Humanitas University of Salford Department of Nuclear Medicine Salford, UK Rozzano, Italy

Contributors

Suzanne Dennan Ann Heathcote Department of Diagnostic Imaging Regional Manger St. James Hospital, Dublin, Ireland Alliance Medical Limited Warwick, UK Thomas Kane Consultant in Radiology & Nuclear Medicine Amy Wareing Victoria Hospital, Blackpool Senior PET CT Technologist and Lancashire PETCT Centre, Preston, UK Preston PET CT Centre Royal Preston Hospital Jean-Marc Vrigneaud Lancashire Teaching Hospitals, UK Medical Physicist Nuclear Medicine Department Peter Julyan Centre Georges-François Leclerc Clinical Scientist Dijon, France North Western Medical Physics The Christie NHS Foundation Trust, UK Sylvianne Prevot Radiation Safety O#cer Katy Szczepura Centre Georges-Francois Leclerc Medical Physics Lecturer Dijon, France Department of Diagnostic Radiography University of Salford Angela Meadows Greater Manchester, UK Unit Manager Preston PET/CT Centre Simona Cola Alliance Medical Technologist of the Department of Nuclear Royal Preston Hospital Medicine Lancashire Teaching Hospitals NHS Trust, UK S. Maria Nuova Hospital Reggio Emilia, Italy Ronald Boellaard Associate Professor and Head of Physics Department of Nuclear Medicine and PET Research VU University Medical Centre Amsterdam, The Netherlands

2 ee ogadAgl edw ...... Imprint ...... Peter . . HoggandAngelaMeadows ...... Chapter 8:Radiographer andtechnologist competencies –education andtraining in PET-CT ...... Simona Cola andPeter Hogg ...... Chapter . 7:Patient care inPET-CT ...... Szczepura . Katy ...... isotopeChapter production 6:PET ...... Peter Julyan . . . . . assuranceChapter 5:Quality control andquality for PET-CT ...... Ann Heathcote, Amy Wareing andAngelaMeadows . . .16 . . . protocol instrumentationChapter andprinciplesofCT optimisation 4:CT ...... Boellaard.Ronald . ..37 ...... imagingChapter instrumentation protocol 3:PET andprinciplesofPET optimisation ...... Jean-Marc Vrigneaud, Sylvianne . Prévot, andPeter AngelaMeadows Hogg ...... Chapter 2:Practical radiation inPET-CT protection ...... andPeter . Thomas Kane Hogg ...... a UKradiologist’s . perspective ...... Chapter 1: The value andlimitations ofPET-CT inroutine clinicalpractice: ...... Peter . . Hogg andGiorgio Testanera . . . . . Preface of andGlossary Terms . andAbbreviations ...... Suzanne Dennan Foreword Contents ...... 99 . . . 3 .69 ...... 76 ...... 4 ...... 81 ...... 7 ...... 5 ...... 88 ...... 54 ...... EANM Foreword Suzanne Dennan

The EANM Technologist Committee was I am grateful for the e(orts and hard work of the established in 1996 to represent European authors, who have ensured the educational val- nuclear medicine technologists and radiogra- ue and quality of this guide. Special thanks are phers within the EANM. Key aims of the Tech- extended to the editors, Professor Peter Hogg nologist Committee include the promotion and Mr. Giorgio Testanera, for their dedication of high professional standards and participa- to the success of this publication. In particular, tion in EANM education and continuing edu- many thanks are due to Siemens Medical for cation initiatives. During the lifetime of the their support and generous sponsorship. committee, the %eld of nuclear medicine has undergone considerable change, particularly It is hoped that this PET-CT book will serve as with the advent of hybrid PET-CT imaging. an invaluable educational tool for all professionals working in PET-CT departments and Since 2004, an annual “Technologists’ Guide” that it will contribute to the quality of their has been produced by the EANM Technolo- daily work. I also hope that this book will ben- gist Committee. This successful series aims e%t those with no or limited PET-CT experi- to develop expertise in key areas of nuclear ence, helping them to start to develop their medicine and to assist with the development understanding of the %eld. I look forward to of high standards of professional practice. the next two books in 2011 and 2012. The current book is dedicated to PET-CT and will be the %rst of a comprehensive three-part Suzanne Dennan series devoted to this important topic. Chair, EANM Technologist Committee

4 Peter HoggandGiorgio Testanera Preface of andGlossary Terms andAbbreviations These outline how PET-CT imaging and ra- and imaging PET-CT how outline These chapters. equipment-related of number a to progresses book the perspective, this duced intro- Having service. routine a providing to is evolving in a with country particular a view PET-CT how about statement clear a it makes because point starting important an is the use of in PET-CT his medical practice. This a radionuclide radiologist’s perspective about quent books. It commences with a chapter on subse- two the for preparation in PET-CT of aspects fundamental some covers book This the chapter. how to extend your knowledge after reading on direction give to or chapter the reading to prior present be should that knowledge background fundamental develop to either ters also have a short reading list, which seeks chap- Finally,most only.references essential list, though we have tried to keep these lists to in 2011 and 2012. Each chapter has a reference next two books in this series will be published about PET-CT, comes at a timely moment. The in mind, this the book, %rst in a series of three in disease certain processes.PET-CT With this of use routine for argument %rm a provide sound %nancial reasons. Taken together, these with combined e#cacy, its to regard in base through a growing and convincing evidence about come has This diseases. prominent several of management and diagnosis the in place its established quickly has and tries rapidly expanding is PET-CT in many coun- many in 5 competence to practice could be achieved be could practice to competence The %nal chapter presents arguments on how eral knowledge about these issues.particular gen- of level existing wehavean anticipated protection, radiation with as and, care, tient pa- on chapter a design’.also ment is There – ‘sta(’,elements ‘patient’three and ‘depart- comprises Thisthis. on chapter substantial a included have we complex, are quirements re- protection radiation PET-CT that Given be conducted to ensure optimal performance. might checks quality what explain they also and work equipment production dionuclide Peter Hogg and Giorgio Testanera Giorgio and Hogg Peter before you begin reading thisbook. insight simple a give to aim that breviations ed below a concise of glossary terms and ab- Finally, as part of this preface, we have includ- petent intheirroles. com- are sta( ensure to seek that processes educational designing also and procedures a valuable resource when conducting PET-CT portantly, we hope that this book will serve as this book and the two related ones. More im- reading enjoy you that hope conducted.We ing to the detail of how PET-CT procedures are (such as beforeQC) radiochemistry progress- issues fundamental more some explore will Building on this book, the second in the series mind whendesigning educationalcurricula. in borne be should considerations what and

EANM Glossary of terms and abbreviation

PET imaging devices and consequently in recent Positron emission tomography (PET) is a to- times there has been a move towards the fu- mographic imaging technique which allows sion of the two medical imaging devices into non-invasive quantitative assessment of bio- one physical unit. The use of one imaging unit chemical and functional processes. A range of to produce two di(erent image datasets has positron emitters are available for use but 18 F become known as hybrid imaging. Examples (combined with FDG – +uorodeoxyglucose) of hybrid imaging devices include SPECT/CT, is the most commonly used. PET-CT has par- PET-CT and PET/MR. ticular value in cancer diagnosis and management but it does have value in many other Molecular imaging pathologies, too. Molecular imaging can be broadly de%ned as the in vivo characterisation and measurement CT of biological processes at the cellular and mo- Computed tomography (CT) is a technique lecular level. Molecular imaging di(ers from that uses an x-ray beam to generate images traditional imaging in that probes (known as that have a very good resolution to demon- biomarkers) are used to help image particu- strate anatomy. lar targets or pathways. Biomarkers interact chemically with their surroundings and in turn PET-CT alter the image according to molecular chang- Integrated PET with CT in a single unit (PET- es occurring within the area of interest. PET CT) has become an established and valued is an excellent molecular imaging modality. imaging modality in clinical routine. Inte- grated PET-CT has been shown to be more Cyclotron accurate for lesion localisation and characteri- A cyclotron is required to generate the pos- sation than either PET or CT alone. PET-CT is itron-emitting radionuclides that are used in an example of hybrid imaging. PET imaging. Ideally cyclotrons are located as close as possible to the PET scanner because Image fusion and hybrid imaging the positron-emitting radionuclides tend to Image fusion involves the bringing together have short half-lives. of two image datasets with the intention of registering them as closely as possible. Gener- ally the two image datasets would have been produced on di(erent types of medical imaging device. Various problems have been incurred when using two geographically remote

6 by radiologists and nuclear medicine physicians.reports of PET-CT cases that had been produced land. This involved him in ‘quality assuring’using mobile scannersclini across the south of Eng- a national audit programme for PET-CT provision clinical PET-CT practice, Tom has been involved 180 toin 200 patients each day. In addition to the a radiotherapy unit which serves approximately geographical area. The teaching hospital also has catchment1.5millionof people overbroada services,suchplasticas andbrain surgery, a to mately 390,000 people. It also provides specialist local residents, serving a population of approxi- hospitalprovides rangeaacute of services to fromthe district general hospital. Thisteaching tonCentre,PET thatapproximatelyis miles15 conducted at a large teaching hospital, the Pres- privatehospital. clinicalHis PET-CTpractice is His private radiology practice is within a nearby hospitallocated theNorthin West England.of eranuclear medicine largeain district general gist.practicesHe radiology andgamma cam- Tom Kane is a dual quali%ed radionuclide radiolo- phasis is placed on cancer detection and staging.what that PET-CT service is used for; particular em England. The second part of the chapter explores mationaboutPET-CT aservice theinnorthof serviceissues andthengives background infor- of this chapter sets out some national UK PET-CT both between and within countries. The %rst part set the context as there can be marked variations PET-CT in routine clinical practice it is important Whenconsidering thevalue and limitations of Introduction Chapter 1: in The valueandlimitationsofPET-CT Thomas Kane andPeterThomas Kane Hogg routine clinical practice: a UK radiologist’s persp cal to - 7 located on hospital premises and it produces is day. 8-hour facility an operates cyclotron A Centre the Generally day. per patients 15 of of scanner time, with a maximum throughput use e#cient for allow which rooms injection three has Centre The syndrome). neoplastic para- of investigation and imaging brain (for category neurology the in referred patients of group small a with (97%), oncology on focussed is workload its of majority The ties. capabili- gating respiratory and cardiac with houses a General Electric 64-slice system CVT 2007, in established Centre, PET Preston The workload isshown inFig.workload 1. its own multi-professional team of sta( and its has Centre vial.The multidose a into put are doses patient 5 than more no generally but number of injections per vial produced varies, day.The per runs radiopharmaceutical two the evidence base alters patient management) cancer; 7, oesophageal cancer; 8, ‘other’ (where neck andhead colorectalsurgery); 6,(post5, cancer; 3, lymphoma; 4, colorectal (pre surgery); pulmonary nodule; 2, non-squamous cell lung Centre,singlecategorised1,tumourtype. by FigurePro%le1: referralsof thePrestonto PET ective

Courtesy of Alliance Medical Ltd, Alliance Medical Lancashire PETCT Centre, Royal Preston Hospital. EANM Within the UK the development of PET services Another catalyst in advancing the availability has been slower than in many other countries. of PET-CT scanners within the UK was the pro- In 2005, a national Framework [1] suggested duction of national guidelines (2005) for lung that research evidence was su#ciently robust cancer management [3]. These guidelines to support the country-wide implementation stipulated that speci%c geographical areas of PET facilities. To help inform resource issues must have rapid access to PET-CT scanning and %nance, this Framework also provided an and that the provision should be a uniformly outline of what would be required to establish high-quality service irrespective of where a and operate a PET scanning service; an ele- patient lives. An interesting departure from ment of future proo%ng was included through the norm within the UK was that a sizeable inferences of where growth areas might lie. proportion of the PET-CT scanners would be On the basis of this Framework, proposals commissioned from private companies; re- were put forward to establish PET imaging imbursement to the private company would services and cyclotrons. Fundamentally this come from taxation, which is in line with the Framework initiated a planned and coordi- principles of the National Health Service. The nated national roll-out of UK PET imaging. An present position of PET-CT within the UK is interesting feature of the Framework was that that a geographically limited routine service PET-CT, and not PET in isolation, was to be the is provided, but the number of PET-CT centres preferred imaging system. is increasing annually in line with the 2005 Framework ambitions. The UK intention was that there would even- Having set the context in which the practice tually be a large number of %xed PET-CT sites sits, let us now explore what that practice with some mobile provision. It was consid- consists of. ered that mobile provision would be important in the early stages; this conclusion was Radiopharmaceutical not reached without good reason. Crowe The most extensively used radiopharmaceuti- [2] gives a clear indication as to why mobile cal in UK clinical practice is 18 F-FDG (+uorine-18 services would be a good starting point for +uorodeoxyglucose), and two factors account establishing a PET service. He indicates that for this. First, the majority of PET-CT scanning mobile PET-CT would distribute the %nancial is conducted to assess malignant disease and risk when entering the market; that mobile there is robust published evidence to support PET-CT would get around the local problem of the use of 18 F-FDG in this context [4, 5, 6, 7, not having readily available trained personnel; 8, 9]; second, because of the general lack of and that one mobile unit could o(er PET-CT to geographical proximity of cyclotrons to PET- several geographical sites on a regular basis. CT scanners, short-lived radioisotopes can-

8 Chapter 1: The value and limitations of PET-CT in routine clinical practice: a UK radiologist’s perspective

not be used easily. 18 F and FDG have good intense activity in excreted urine in the renal characteristics, e.g. high spatial resolution, a collecting systems and bladder. Lower grade relatively long half-life that permits longer syn- activity due to glycolysis is shown in the liver, thetic procedures to be engaged with (label- and there is similar activity in bone marrow. ling) and su#cient activity in a sample for the The lungs, although having low-level meta- distance between the cyclotron and the PET bolic activity, are apparently totally free of scanner to be quite large. FDG is a modi%ed activity due to over-correction from the very glucose molecule that is taken up into cells low CT value of air-containing lung in the at- by the same pathway as glucose, usually via tenuation correction process. the GLUT-3 transporter, but it cannot be me- tabolised beyond initial phosphorylation to Figure 2: Normal biodistribution of 18 F-FDG FDG-6-phosphate. FDG uptake therefore de- in an adult. From left to right, images are: EANM picts normal and abnormal metabolic activity. coronal CT, coronal PET, fused PET-CT and MIP PET Figure 2 illustrates normal biodistribution of 18 F-FDG in an adult. Normal metabolic activity is minimised in clinical imaging by asking patients to starve for a period of hours prior to injection, with the intention of achieving a hy- poglycaemic state (low blood sugar). There is normally a high uptake within the brain, which has obligatory glucose metabolism. There is also high activity in the urinary system as FDG is not trapped by the renal tubular cells, unlike normal glucose. There is usually low-grade

FDG activity in the liver due to glycolysis and Courtesy of Alliance Medical Ltd, PETCTAlliance Medical Lancashire Centre, Hospital. Royal Preston there can be variable uptake in myocardial cells. Muscle that has recently been exercised PET-CT in routine practice will also show FDG uptake and for this reason Within the Preston PET Centre, FDG is used patients are encouraged to limit physical activ- almost exclusively for cancer detection, stag- ity prior to injection and scanning. ing and recurrence. This is consistent with PET- CT practice elsewhere; world-wide, around Figure 2 demonstrates intense uptake in the 90% of PET-CT clinical workloads generally grey matter and basal ganglia of the brain, are for cancer detection. In the context of moderately intense myocardial uptake, and cancer staging, PET-CT is most often used to

9 target potentially curative therapy, often by For the purpose of the rest of this chapter we reducing the frequency of unnecessary sur- shall consider only two cancers that are diag- gery. This is particularly true for lung cancer, nosed and managed within the Preston PET where metastatic spread may occur remotely Centre – non-small cell lung cancer (NSCLC) without more local tumour spread. The whole- and colorectal cancer. body ability of PET-CT, combined with its high sensitivity, is therefore of considerable clinical Non-small cell lung cancer value in comparison with other contemporary Of all the cancers, lung cancer is the most imaging modalities. frequent and the most lethal. Predisposing factors for this cancer include smoking and Until the Preston PET Centre opened as a mo- passive smoking, asbestos, air pollution and bile service in 2005, and subsequently as a exposure to radon. On occasion lung cancer %xed site in 2007, cancer imaging depended has no symptoms; on other occasions the upon other radiological techniques, such as symptoms can be varied and include: per- magnetic resonance (MR), ultrasound (US), sistent cough, dyspnoea, chest discomfort, computed tomography (CT) and gamma haemoptysis, hoarseness, anorexia and weight camera nuclear medicine (NM). For many loss for no known reason. common cancers the accuracy of CT staging can be limited; MR, US and gamma camera NSCLC comprises several cancers (e.g. ade- NM also have limitations. Between 2005 and nocarcinoma, squamous cell carcinoma and 2006 Tom’s perspective on the use of medical large cell carcinoma) and each has its own imaging for the evaluation of cancer changed characteristics. For most patients with NSCLC, because the increased speci%city and sensi- current treatments do not cure the cancer. tivity of PET-CT, compared with conventional NSCLC can be diagnosed or excluded by imaging alternatives, became increasingly ap- using a broad range of diagnostic tools and parent. The role of PET-CT was also becoming techniques, including physical examination, apparent to physicians and surgeons through chest x-ray, CT, sputum cytology, %ne-needle the peer-reviewed journals within their own aspiration, bronchoscopy and PET-CT. Nearly specialities. This led to pressure for the provi- all types of NSCLC are FDG avid. Figure 3 illus- sion and use of PET-CT in clinical pathways for trates an FDG scan on a patient with NSCLC. common cancers within the newly opened Using the TNM staging classi%cation [10], this Preston PET Centre. patient has T2, N0, M0. As can be seen from the scan, there is intense FDG update in a small hilar tumour with no abnormal FDG uptake elsewhere.

10 Chapter 1: The value and limitations of PET-CT in routine clinical practice: a UK radiologist’s perspective

Figure 4: 18 F-FDG scan on another patient with NSCLC, showing multiple tumour deposits in bone and a tumour nodule adjacent to the left kidney Courtesy of Alliance Medical Ltd, PETCTAlliance Medical Lancashire Centre, Hospital. Royal Preston EANM Figure 3: 18 F-FDG scan on a patient with NSCLC; note the intense uptake in the tumour

By contrast, Fig. 4 shows a patient with an extensive lung tumour, with FDG uptake in Courtesy of Alliance Medical Ltd, PETCTAlliance Medical Lancashire Centre, Hospital. Royal Preston infected lung around this. Conventional imaging had suggested this might be amenable to For the solitary lung nodule, PET-CT has been radical lung resection, but PET-CT revealed shown to be highly accurate at di(erentiat- multiple tumour deposits in bone, seen on ing benign from malignant disease [11]; this these images in the spine and bony pelvis as approach has a particular value when CT and well as the left humerus, and a tumour nodule chest radiography alone are indeterminate. in the abdomen adjacent to the left kidney. PET-CT also has value where biopsy is risky.

FDG PET-CT has a very low false negative rate for both NSCLC and small cell lung cancer, although this rate increases signi%cantly in small lesions. Most centres regard 7 mm as the lower limit for reliable exclusion of FDG uptake, and for this reason most published guidelines use 10 mm as the lower size limit for FDG assessment. However, for FDG-negative lesions above 10 mm, most authors would support a policy of observation using serial

11 CT. Management of FDG-positive lesions is In proven NSCLC the prognosis is strongly re- complicated by potential false positive %nd- lated to tumour staging; in patients in whom ings; FDG uptake is not inevitably due to curative surgical resection or radical radio- malignancy, and in particular granulomatous therapy is not possible, the 5-year survival is infection (e.g. tuberculosis or, in North Ameri- e(ectively zero. In the British population fewer can populations, histoplasmosis) can give very than 20% of cases of NSCLC will present at high FDG uptake. Again, most authors would stage 1 or 2a, where curative surgery may be recommend either excision or percutaneous considered. Conventional imaging assesses biopsy, depending on the patient’s general nodal stage by size criteria; however, in the state of health and the likelihood of curative mediastinum, size can be a poor predictor of treatment. Figure 5 illustrates an example of tumour spread due to the presence of reactive a patient presenting with a chest mass which nodal enlargement. FDG has been shown to all imaging modalities (including PET-CT) sug- have a greater predictive value in assessing gested to be a small lung cancer, but which mediastinal nodes, though there are a num- was found following surgery to be a focal tu- ber of publications showing that FDG uptake berculous infection. can vary, again due to reactive change. A local study in the Preston PET Centre con%rmed Fig. 5. Imaging results in a patient with a focal an approximately 8% error rate, both for FDG tuberculous infection mimicking lung cancer uptake in benign nodes and for FDG-negative involved nodes. This has led to current advice being to surgically sample FDG-negative en- larged nodes, and FDG-positive nodes that are not unequivocally malignant, prior to a treatment decision. FDG has shown a clear ability to detect occult metastatic disease, particularly in the skeleton, over conventional CT staging. Currently the staging of NSCLC forms the single largest referral group in the Preston PET Centre, at 45% of referrals. Courtesy of Alliance Medical Ltd, PETCTAlliance Medical Lancashire Centre, Hospital. Royal Preston

12 Chapter 1: The value and limitations of PET-CT in routine clinical practice: a UK radiologist’s perspective

Colorectal cancer In the Preston PET Centre the commonest re- Colorectal cancer is the second most preva- ferral indication in colorectal cancer is to stage lent cause of death by cancer in developed patients whose primary tumour has been countries. Symptoms include rectal bleeding, resected and who are subsequently being change in bowel habit, an abdominal lump, assessed for suitability for either liver or lung weight loss and anaemia. Colon cancer can be resection for presumed localised metastatic diagnosed or excluded by using a broad range disease. As in other cancers, the exclusion of of diagnostic tools and techniques, including other sites of disease is considered highly pre- barium enema, colonoscopy, CT colonoscopy, dictive of curative success for localised resec- blood tests, US, MR and PET-CT [12]. Anatomi- tion. Whilst PET-CT has been shown to have cal methods of imaging colorectal cancer a lower sensitivity for liver lesions compared have been found to be suboptimal, not least with MR, and a lower speci%city as compared EANM because microscopic spread reduces cure with CT portography, the ability of PET-CT to rates and late diagnosis by anatomical meth- detect metastatic disease at other sites out- ods confounds this. PET-CT can be valuable for side the liver or lung has given this technique assessing the extent of metastatic disease and an overall greater predictive power for surgical as such it is useful in di(erentiating patients planning. Figure 6 illustrates a patient who in whom surgery is particularly valuable, i.e. presented with colon cancer. Conventional those with locally resectable disease. imaging had indicated a solitary metastatic deposit in the liver (the large FDG-avid lesion), PET-CT is not generally used for primary and liver surgery was being considered. FDG detection or screening but it does have par- PET-CT, however, showed additional meta- ticular value in identifying loco-regional and bolically active nodes in the liver hilum and disseminated disease. It can also be used to adjacent to the porta hepatis, indicating in- detect tumour recurrence when blood tests operable tumour. [e.g. carcino-embryonic antigen (CEA)] raise suspicion and when other radiological tests are normal. Additionally, PET-CT has value in assessing CT-detected abnormalities, particularly soft tissue thickening at the site of previous surgical resection. The presence or absence of FDG uptake has a high predictive value in distinguishing surgical scarring from tumour recurrence, though evaluation can be hampered in the presence of infection.

13 Fig. 6a A liver deposit (large lesion) and a small How will the PET-CT service develop? liver hilum node are visible in a patient with For the Preston PET Centre there is a relatively colon cancer. Fig. 6b Further FDG-avid nodes low referral rate of lymphoma cases, which for the UK PET-CT practice generally form the 6a single largest patient group. This low referral pattern is highly likely to change in time and there should be an increasing trend to image this patient group. As elsewhere in the UK, there is an increasing number of referrals to assess the potential for secondary deposits from oesophageal cancer that are being considered for surgery or radical therapy. There is also an increasing trend for imaging pathologies which fall outside the currently agreed referral pathways; these include malignant mela- noma, gynaecological cancers and nodal staging of cervical carcinoma. This also re+ects UK practice, where recent assessment by the UK PET-CT Advisory Board shows clinical practice 6b evolving as PET-CT becomes a more mature and established medical imaging technology. 6a + 6b: Courtesy of Alliance Medical Ltd, PETCTAlliance Medical Lancashire Centre, Hospital. Royal Preston

14 PET Clinics2008;3:147-153.PET Med 2004;34:180-97.Med 8. Delbeke D, Martin D, Delbeke W.Martin 8. reality clinical a PET-CT and PET Making P. Crowe 2. Guidance/DH_4121029 cationsandstatistics/Publications/PublicationsPolicyAnd- Crown. 2005, Health of Department England. in Services Tomography(PET) sion Emis- Positron of Development the for Framework A 1. References solitary pulmonary nodules pulmonary solitary solid partly or solid of characterization the in PET-CT of Jeong S, 11. Lee K, Shin K, Bae Y, Kim B, Choe B, et al. 2009 www.cancer.gov/cancertopics/factsheet/Detection/sta emission tomography in breast cancer imaging Eubank W, Manko( D. 9. stomach Dehdashti F, Siegel B. 7. and neck cancer, including thyroid carcinoma Schöder H, Yeung H. 6. 2004;34:166-79. mography in the evaluation of lymphomaR. Bar-Shalom Z, Keidar O, Israel 5. 2005;60:1143–55.major advance. ClinRadiol Wechalekar K, Sharma B, Cook G. PET-CT in oncology – a 4. media/pdf/CG024niceguideline.pdf and treatment of lung cancer. 2005. forExcellence. Clinical Institute National diagnosisThe 3. cott Williams and 2003.Wilkins; Clinical molecular anatomic imaging. Philadelphia: Lippin- editor.G, Schulthess VonIn: services. PET mobile through 12. Lonneux M. Lonneux 12. Sheet.Staging: questions answers.and Available at National10.Cancer Institute. National Cancer Inst Med 2004;34:224-40. colorectal carcinoma References Chapter 1 . Semin Nucl Med 2004;34:198-208. NuclMed . Semin FDG-PET and PET-CT in colorectal cancer colorectal in PET-CT and FDG-PET . Semin NuclMed. 2004;34:209-23.. Semin Positron emission imaging of head Current and future uses of positron http://www.dh.gov.uk/en/Publi- Neoplasms of the esophagus and PET and PET-CT forof evaluation PET-CT and PET . Lung Cancer 2008;61:186-94. www.nice.org.uk/nice- Positron emission to- emission Positron . Semin Nucl Med . Semin Nucl . Semin Nucl ituteFact E#cacy E#cacy : http:// ging . , 15 ging. Philadelphia: Lippincott Williams and 2003.Wilkins; Von Schulthess G, editor. Clinical molecular anatomic ima- 2004. imaging in oncology. Berlin Heidelberg New York: Springer; Czernin J, Dahlbom M, Rabit O, Schiepers C. Atlas of PET-CT 2007.Elsevier; – techniques CT and technology. 6th ed. St. Louis: Mosby/ P, Christian PET-and medicine Nuclear Waterstram-RichK. 2006. Arnold; Hodder London: ed. 2nd tomography. emission Barrington SF, Maisey MN, Wahl RL. Atlas of clinical positron Suggested reading

EANM Chapter 2: Practical radiation protection in PET-CT Jean-Marc Vrigneaud, Sylviane Prévot, Angela Meadows and Peter Hogg

a) THE PATIENT Optimisation of administered activity Jean-Marc Vrigneaud Radioactive doses should be as low as reasonably achievable but high enough to obtain the Introduction desired diagnostic information. The radiation Radiation exposure to the patient from a PET- dose to the patient may depend on factors CT scan is both external, from the CT scan, and associated with the PET scanner, in that these internal, from the injected PET radiotracer. Ac- factors in+uence the amount of administered cordingly, the radiation burden to the patient activity required to provide an image quality can be relatively large, with a total e(ective good enough to produce a reliable diagnosis dose of 25 mSv commonly cited in the case (see next section). of PET-CT studies [1]. Dose reduction can be achieved by careful attention to the CT imag- 18 F-FDG is by far the most commonly used ra- ing parameters and the administered activity diopharmaceutical in PET. The radiation dose of the PET tracer; this is particularly important from 18 F-FDG can be calculated using tables in children. This review will outline the factors from International Commission of Radiation that a(ect radiation dose in both modalities Protection (ICRP) Publication 80 [2]. These and will try to give some insights into dose tables provide dose data, using ICRP publica- reduction in PET-CT imaging. tion 60 dosimetry, in the case of standardised individuals (children aged 1, 5, 10 and 15 years PET dosimetry and adults). Accordingly, these data do not Considering the intention of optimisation in account for di(erences among individuals in radiation protection, the aim is to achieve the terms of their pharmacokinetics and should minimum patient radiation dose consistent not be used to evaluate the risk to a given with diagnostically acceptable image quality. individual. However, they do provide generic In nuclear medicine, this can be done primar- assessments of organ and e(ective doses that ily by: are su#cient to permit comparison of di(erent techniques or di(erent medical examina- t Appropriate selection of the best available tions. For instance, in the case of 18 F-FDG, the radiopharmaceutical and its activity, with adult dose per unit activity administered is 19 special requirements for children µSv/MBq. This leads to an e(ective dose of almost 7 mSv for a typical administered activity t Appropriate image acquisition and pro- of 370 MBq. The organ receiving the highest cessing absorbed radiation dose is the bladder.

16 tration of the lowest possible activity in chil- in activity possible lowest the of tration adminis- to paid be should attention Careful spect to carcinogenesis. At our institution, our At carcinogenesis. to spect re- with radiosensitive more be to known are they as patients paediatric to drawn be unusually high or low. Special attention should are activity administered of levels the where situations identify to used be should levels reference diagnostic possible, Whenever this will cause unnecessary radiation exposure. necessary, is examination the of repetition if diagnostic value. It must be borne in mind that of maintenance with consistent is that dren dosage card [3]. ciation of Nuclear Medicine (EANM) paediatric weight, according to the new European Asso- patient of function a as administeredactivity surface area). For example, Table 1 summarises body (e.g. criteria appropriate to according proaches exist to balance the paediatric dose ap- Other section). next (see equipment the on depending adopted be may level dose inject 5 MBq/kg of the standard protocol for clinicalstudiesisto Chapter radiation 2:Practical inPET-CT protection 2 1 (3D acquisitions)[3] from dose Radiation 1: Table Estimated from ICRPpublication80[2] E(ective dose(mSv) 18 egt(g 01 25 70 55 32 19 10 Age (yr) Weight (kg) Estimated from anatomical data -D diitrdatvt Mq 86 0 6 196 163 102 65 38 (MBq) F-FDG administered activity 1 18 2 F-FDG but an even lower 18 F-FDG according to the new EANM paediatric dosage card dosage paediatric EANM new the to according F-FDG . . . . 3.7 4.0 3.8 3.3 3.6 01 Adult 15 10 5 1 17 theresigni%cantis noise, resolutionthen and curacy of the tracer concentration and noise. If +uenced by spatial resolution, quantitative ac- in- also is quality image Basically, [4]. PET PET in ratio noise to signal the to related metric a essentially is which (NECR), rate count lent tered activity as a function of the noise equiva- attempts have been made to adjust adminis- metrics exist to de%ne image quality. Recently, very di#cult task in PET imaging as no simple a is performance optimal of Determination cessing pro- and acquisition image of Optimisation keep still during theexamination. stillduring keep foring time and the possibility the patient to compromise may be required between imag- if image quality is good enough. Ultimately, a dose injected reduced or time scan reduced In contrast, smaller patients may bene%t from fractions. random and scatter increased and er the noise because of increased attenuation tient motion. The heavier the patient, the high- pa- no and image the within level noise low a require acquisitions PET lost. are accuracy

EANM Considering PET scanners, acquisition and CT dosimetry reconstruction parameters should be cho- CT delivers a relatively high radiation dose to sen such that the image quality is optimum. the patient. Its use has increased rapidly as a For example, in paediatric studies where the result of the tremendous advances in CT tech- radiation burden can be an issue, 3D acqui- nology that make it extremely user-friendly for sition mode should be used because of the both the patient and healthcare workers. In enhanced sensitivity of the PET scanner in 3D Western countries, though it represents only mode compared with 2D mode. Reconstruc- a small fraction of all medical procedures in- tion parameters should also be optimised as volving ionising radiation, its contribution to a function of the region being scanned. In the collective e(ective dose is quite large (up general, iterative algorithms are the standard to one-third). in PET because of the relatively high statisti- cal noise originating from the emission data. Overview of factors a"ecting radiation In the most recent implementation of these dose in CT algorithms, better image quality is achieved by The factors a(ecting radiation dose can be including all the corrections needed (geom- classi%ed as intrinsic and extrinsic. Intrinsic fac- etry, normalisation, dead time, scatter, attenu- tors are related to the geometry and design of ation, random events) in the iteration loop. the scanner (tube, focus, collimator, %ltration, detector design, etc.) and cannot be modi%ed New technologies will continue to improve by the user. Extrinsic factors are those param- image quality, providing the physician with eters that can be adjusted by the user, and the opportunity to reduce scan time per bed it is these parameters that mainly determine position or to lower the injected dose. For in- the patient dose. Optimisation of CT radiation stance, time-of-+ight (TOF) imaging seems to dose is rather challenging and requires a good be particularly promising in terms of achieving knowledge of how the various factors a(ect a lower level of noise and better resolution the absorbed radiation dose; these concerns compared with non-TOF imaging. This bene%t are summarised in Table 2. has already been shown in large patients, with the TOF gain being more signi%cant when limited acquisition time and high attenuation reduce the total counts in the image [5]. In the same way, dynamic studies with low statistics should also bene%t from the enhanced image quality achieved with TOF imaging.

18 Chapter radiation 2:Practical inPET-CT protection Table 2:User-adjustable a(ecting radiationdoseinCT factors For example, in the case of paediatric patients, shape or attenuation than the CTDI phantom. dose for objects of substantially di(erent size, CTDIThe with diagnostic reference levels. patient dose or to compare the radiation dose study the in+uence of technical parameters on standardised CTDI phantom. It can be used to a for volume scan the within dose average the represents It initiation. scan to prior sole con- user the on displayed is mGy) in Index CTDIThe quantities dosimetric Displayed CT kVp mAs Time perrotation mA Patient size Slice thickness Collimation Scan lengthScan Pitch vol vol does not represent the average the represent not does (Computed Tomography Dose Dose Tomography(Computed proportional to (kVp)proportional X-rayThe tubevoltage measured peak. outputisapproximately inkilovolt ducing themAs reduces theradiationdosebutincreases thenoiseby (1/√mAs). The product ofthetubecurrent andtime. related Linearly to radiationdose. Re- The exposure timeperrotation related islinearly to radiationdose. The related tubecurrent islinearly to radiationdose. tion dosethanthelarger patient. With thesameacquisitionparameters, thesmallerpatientreceives ahigherradia- image. to 1/√(slicethickness).The noiseisproportional This parameter a(ects thepatientdoseindirectly by governing thenoisein atthenarrow beamcollimationrequiredine#ciency for narrow slicewidths. to systems have have detector-row aradiationdose Multiple beenobserved CT tient irradiated increases because of the exposure larger scan region but the e(ective dose increases The average radiation dose within the volume may va fective mAs ormAs/slice =truemAs/pitch. the e(ective mAs ormAs/slice isused, radiation doseisuna(ected by pitch. Ef- increased aspitch isincreased to maintainimagenoise. As aconsequence, when pitch. However, onmultipledetector-row CT, tubecurrent isautomatically With respect to otherexposure parameters, dose isinversely to proportional voltage from 140to 120kVreduces patientdoseby 40%. 2 . With respect to otherexposure parameters, changing the 19 neck, thorax,abdomenorpelvis). neck, (head, scanned being region the of function using appropriate conversion coe#cients as a fective dose can be estimated from the DLP [6] account the length of the scan acquisition. Ef- to the complete scan acquisition. It takes into the total energy absorbed that is attributable re+ects DLP The quantity. displayed another is mGy·cm) in (DLP product dose-length The appropriate phantoms adapted to childsize. of use the requires applications paediatric in CTDIof determination Accurate 2. of tor CTDIthe vol can be underestimated by a fac- a by underestimated be can of additional tissues and organs. linearly as the length of the pa- ry slightlyry when using a

vol EANM Dose reduction in CT patient will need a higher than average mAs to As in PET, patient dose in CT is interconnected counteract the e(ect of increased attenuation. with image quality (spatial resolution, noise, Reducing the scanning length and minimising slice thickness). For example, to reduce the level the number of scans in an imaging study are of noise by a factor of 2 with respect to spatial also helpful in optimising patient dose. resolution and slice thickness, it is necessary to increase the dose fourfold. When optimising On modern CT scans, automatic exposure patient dose, this trade-o( should always be control (AEC) systems adjust the x-ray tube kept in mind and the CT parameters should current (mA) in real time during gantry rota- always be adapted according to the contrast tion in response to variations in x-ray intensity needed in a given region. Protocols should at the detector. AEC systems enable the CT also take the individual patient into account user to prescribe a measure related to image by selecting parameters according to patient quality so as to modulate the tube current as size, age and gender and the clinical question. a function of size, shape and geometry of the region being scanned. The main advantage of Amongst all parameters listed in Table 2, the this technique is the consistent image quality tube current (mA) is often the least stan- obtained within a patient as the patient’s at- dardised. Tube voltage and gantry rotation time tenuation varies but also from one patient to are generally %xed for a given clinical applica- another, irrespective of the patient size. These tion. Lower tube voltages can be used for small systems should be used with caution to en- adult patients or children or in speci%c proto- sure that the required image quality is always cols that require a low radiation dose (e.g. in speci%ed appropriately. To this end, the dis- pregnant woman). Reducing the tube voltage played dosimetric quantities can be checked improves image contrast but also reduces the to identify any misuse of these systems. penetration of x-rays and increases the noise in the image. The fastest rotation time should be PET-CT dosimetry used to minimise motion blurring and artefacts. PET-CT dosimetry will depend strongly on imag- Ultimately, the CT operator should take patient ing protocols. In most cases, patients referred for size into account when selecting the mA (or a PET-CT procedure do not require stand-alone mAs). Several technique charts exist to adjust diagnostic CT or contrast-enhanced CT. High- mA according to various criteria (e.g. patient quality anatomical details are not essential and weight or patient thickness) [7]. For example, a higher level of noise can be tolerated in the for body CT imaging, a reduction in mA by a images. Also, there is no need to discriminate be- factor of 4–5 from adult techniques is accept- tween various kinds of soft tissue. Indeed, the CT able in infants [8]. On the other hand, a large component is used mostly for anatomical corre-

20 andlocalisation, reductiona parameters CT in componentCT usedisonly for co-registration agesrequired radiologyin departments. theIf reduction compared with diagnostic quality im- This means that there is a great potential forlation dose of the PET data and attenuation correction. 10x1.5 mm, pitch 0.7, scan length 1 m (CT-Expo 1.7, Simulationswere carried outfor GEMINI a GXLPET-CT Table3:Radiation exposure from whole-bodya sc CT theradiation exposure without compromising tefacts. In this case, it is possible to further withre linearity maintained and without any CT ar- curate representation of the attenuating tissue, Here, the only requirements are to obtain an ac- relevant if only attenuation correction is needed. on the patient weight [9]. This trend is even themore mAs can be as low as 10-40 mAs depending qualityfor anatomical correlation. For example, is possible while maintaining acceptable image Chapter radiation 2:Practical inPET-CT protection dose due to the the we%rst part, learnt that the internal adult FrommSv. 2-15 of range the in be may scan CT the from dose e(ective the patient, adult an for 3, Tableto According injected. tracer PET of amount the extent, lesser a to and, protocols, the number of CT scans performed imaging the on depend will studies PET-CT As a consequence, the patient exposure from (cieds mv 215 12 8 4 2 E(ective dose(mSv) L myc)172357801100 880 587 293 147 DLP (mGy·cm) CTDI vol (mGy) 18 F-FDG tracer is approximate- . . . . 10.7 8.6 5.7 2.9 1.4 04 010150 120 80 40 20 duce 21 © G. Stamm, Hannover and H.D. Nagel, Hamburg) scan operating at di(erent mAs per slice. In somee(ective dose obtained from awhole-body CT arenotoptimised. Table shows3 examples of dose can be relatively high if the CT parameters the mid thighs to the eyes. The resultant e(ectivewhole-body imaging is to scan each patient from extentscan.theCT ofThe current method for Another speci%city of PET-CT imaging is the axial to 80-90 kVp in small adult patients or children. as 10 mA and the tube voltage can be reduced siondata[10]. The tubecurrent lowascanbe thequality theattenuation-correctedof emis- achievable. and CT to maintain doses as low as reasonably atric adjustments should be used in both PET paedi- possible, Whenever adults. to relative and they also have a potential for a longer life to be moreknown radiosensitive than adults on the imaging protocol chosen. Children are adult patient will thus be 9-22 mSv depending the for burden radiation total mSv.The 7 ly according to the clinical indication. approach and adapt scan length and scan quality cases,mayitmorebe appropriate tailorto this anfor PET-CT studies functionasa ofmAs. scanner (10-sliceoperating CT) at120kV, mAs/slice

EANM b) THE STAFF Sta( doses must be maintained as low as pos- Sylviane Prévot sible. National regulation limits are applied as part of the control of practice (Table 4). Introduction Operational protection will be based on the Compared with the radiation exposure of sta( following: traditionally used to handling 99m Tc and other low-energy radiopharmaceuticals, the radia- t Prior risk assessment and optimisation of tion doses reported from PET are much high- protection in all workplaces er [11, 12, 13]. In the last few years the rapid expansion of PET-CT facilities and the intro- t Delineation of areas (controlled/super- duction of positron emitters in conventional vised); avoidance of any accidental en- nuclear medicine departments have given rise trance to new radiation safety concerns for radiographers and nuclear medicine technologists. As t Classi%cation of workers into two catego- a consequence, working practices have had ries (A/B) according to the doses likely to to be reviewed and sometimes modi%ed in be incurred in normal working conditions order to minimise individual and collective radiation exposures. t Information and training of exposed workers

General principles of radiation protection t Appropriate measures to protect pregnant The principle of optimisation [ALARA (as low as sta( from the hazards of ionising radiation reasonably achievable) philosophy] has been part of the European basic safety standards t Monitoring of exposures and adequate since the 1980s. It was re-emphasised in the medical surveillance Euratom Council Directive 96/29 [14], which had its roots in ICRP 60 [15]. In the context of the t Monitoring of the working environment optimisation of occupational exposure, radiation safety issues need to be addressed before PET tracers are handled. Source-related restric- tions on the prospective doses to individuals in planned situations – dose constraints – must be used when designing new premises. The layout of the department must be considered as well. This careful approach aims to ensure the safe practice of PET-CT imaging.

22 avoids scattered radiation out of the scanner the of out scatteredavoidsradiation doors and walls room.Adequatein shielding a high risk of external exposure for sta( in the scattered radiation from the patient results in scan, CT the During beam. x-ray an uses CT studiesComputed tomography (CT) safetyFDG PET-CT issuesin Radiation Chapter radiation 2:Practical inPET-CT protection Table limitsaccording 4: Dose to Euratom Council Directive 96-29[14] HVL, half-value layer apart. oppositedirections,emittedin511 keV, 180° convertedis intotwo annihilation photons of electronfree(e a collisionmatterwithbeforecombineswithit a few millimetres, losing kinetic energy through nucleus undergoing β Positrons(e Physical of characteristics 18 Table 5:Attenuation ofx-rays inradiology over aged 18years or students apprentices & Exposed workers, Fetus eea ulc1mv1 S 50mSv·cm 15mSv 1mSv public General F-+uorodeoxyglucose (FDG) PET imagingF-+uorodeoxyglucose (FDG)PET Voltage 18 (kV) F emissions are shown in Table 6. 5 .66 4<5×10 <1×10 74 82 61 70 0.26 0.2 150 100 0019 7<1×10 97 91 0.1 50 + ) are emitted from a proton-richemittedafrom are ) - ). The mass of the ethe of mass).The "cieds E Equivalent dose(H) E"ective dose(E) + 20 mSv peryear (average over a Max. 50mSv in Max. decay. A positron travels 5-year period) hl oyLn feeSi Extremities Skin Lens ofeye Whole body 1 mSv over pregnancy (mm Pb) 1 year 18 HVL F + and eand Attenuation (%) 0.35 mmPb 5 S 500mSv·cm 150 mSv -

23 are isolated from the source/patient within the room. In normal working conditions, operators datory (Table 5). (Table datory man- is x-rays of attenuation 74% than more providing apron lead 0.5-mm a case that In someone beside the patient may be justi%ed. of presence the Occasionally room. control Table 6: 15114649 3 1 97 634 194 511 E1 (keV) 18 Attenuation (%) Gamma E F emissions(Delacroix etal. [16]) 0.5 mmPb % Beta (EBeta (keV) -2 -2 E max Transmission (%) % ) 2 mmPb 500 mSv (keV) Electron E -3 -3 -6 % EANM Risk of internal exposure 18 F-FDG in a vial or syringe is a source emitting External contamination of skin or internal con- high-energy gamma photons and beta particles. tamination by inhalation or ingestion may re- Despite a high dose rate, 0.63-MeV positrons sult in internal exposure. 18 F has a short half-life have ranges of 0.9 mm in glass and 1.7 mm in (T = 110 min) and radiotoxicity is moderate; PMMA. Most of them are stopped in the walls of 1/2 consequently the risk of internal exposure is a vial. Exposure rates at short distances without a relatively low despite β emission. shield can be extremely high: more than 1 Sv·h -1 in contact with a syringe (Table 7). The dose limit Risk of external exposure for extremities can be reached in only 21 min With a gamma ray constant in air being ap- when holding 500 MBq of 18 F in an unshielded proximately 10 times higher than the 140-keV syringe, whereas it takes 3 h with 99m Tc (Table 8). emissions from 99m Tc, 511-keV annihilation photons result in a high and sometimes underestimated risk of external exposure for sta(. Table 7: Exposure rates from 18 F without a shield (Delacroix et al. [16]) Exposure rate (mSv·h-1) On contact At 10 cm At 40 cm Multidose, 10-mL vial, 6 GBq 4200 96 6 Single dose, 10-mL vial, 500 MBq 350 8 0.5 Syringe, 5 mL, 500 MBq 1450 - -

Table 8: Exposure rates when holding various radionuclides in an unshielded syringe Exposure rate for Time required to Dose rate on contact Radionuclide 500 MBq reach dose limit (µSv·h -1 ·Bq -1 a) (mSv·h -1 ) (500 mSv) 99m Tc 3.5×10 -4 174 2.9 h 131 I 1.1×10 -3 550 54 min 18 F 2.9×10 -3 1450 21 min aDelacroix et al. [16] Following administration, beta particles are men is 0.26±0.04 µSv·h -1 ·MBq -1 immediately stopped in soft tissues. The patient is a moving after injection and 0.15±0.04 µSv·h -1 ·MBq -1 1 source emitting scattered 511-keV photons. h later, just before installation after emptying The mean dose rate measured with a Babyline the bladder (Table 9). 81 (Nardeux) at 0.5 m from the patient’s abdo-

24 Table 9:Exposure rates measured andat0.5mfrom oncontact apatient Chapter radiation 2:Practical inPET-CT protection HDP, hydroxymethylene diphosphonate tices, including: prac- working local and department the of in every PET-CT facility according to the layout implemented be must strategy protection A Practical steps to control external exposure tained ataminimumlevel. exposuresthat so reducedbe can main- and developed be to have procedures handling and devices shielding speci%c Therefore, t t t nal exposure: must be considered in terms of exter- critical situations procedures.Three medicine clear nu- standard in involved those than higher much be can fromshield a without patient a holding 1 when 8 incurred be to likely doses the that demonstrate clearly %gures These Radiopharmaceutical 18 18 99m 99m F-FDG sources or when at a short distance short a at when or sources F-FDG -D 30Mq ntlain(1h 6 52 91 364 590 installation(+1h) F-FDG (350MBq) followingF-FDG (350MBq) injection cHP(2 B)isalto + )4 9 24 40 180 Tc-HDP installation(+3h) (922MBq) Tc-HDP following (470MBq) injection Positioning the patient on the scanning bed Injecting thepatientInjecting dispensingDose 25 t t t t t Dose dispensing is performed inthehotlab- dispensingisperformed Dose t room. control scanner the is as areas, supervised as study. The rest of the time they are designated CT the during and/or present is patient tive room are controlled areas as long as a radioac- area. Injecting/resting rooms and the scanner controlled a as designated is which oratory, Adherence to well-established standard well-established to Adherence with accesstorestricted properly trainedsta( areas of delineation Appropriate environment working the of surveillance Radiological ofexposuresClose monitoring Close attention to sta(training time,principles: distance, shielding Use of the three basic radiation protection inradiationsafety practice best maintain to procedures and policies ncnatAt 0.5m contact On Exposure rate (µSv·h -1 ) EANM Optimisation of sta( exposures Maximising the source–operator distance Absorbed dose – average dose over a tissue or t Make use of the inverse square law an organ – is a function of the dose rate and of the time spent near the source of radia- t Use long tongs (25-40 cm) to place and tion. In well-managed operations, protection remove unshielded vials in the dose cali- bene%t involves a balance between several brator factors so that dose rates can be signi%cantly reduced without a corresponding substantial t Draw up with a spinal needle (20G×90 mm) increase in time. t Use a trolley to carry doses from the hot Minimising time laboratory to the injecting room t Prepare every process very carefully and perform all radioactivity tasks as swiftly as t Avoid staying beside the patient unneces- possible sarily after injection t Check volume required before drawing up t Use the intercom to communicate with then dispense dose as rapidly as reason- patients able t Use remote viewing to oversee patients in t Conduct all patient examinations, give the resting area/scanning room clear explanations and allow time for questions before FDG is injected t Direct patients rather than escort them unless they need support t Optimise the injection procedure gaining a good IV access (e.g. cannula with a three- The mean dose rate measured at the patient’s way tap) before handling the activity abdomen just before installation is 2.5 times lower at 1 m than at 50 cm. t Spend only as long as necessary when positioning patients

Experienced and well-trained radiographers and nuclear medicine technologists generally perform manipulations more rapidly. Sta( rotation also contributes to reducing the time of individual exposure.

26 Courtesy of Centre Leclerc. t paredwith in lead (Pb) and 2.7 mm in tungsten Com- (W). thickness resulting in 50% attenuation, is 4 mm Figure 1:Hotcell(Lemer-Pax) t Attenuation is a stochastic process depending on:and theoperator to beprotected Placing adequate shields between the source Chapter radiation 2:Practical inPET-CT protection when handling greater.Therefore, thicker shields arerequired 18 layer(HVL) of 511-keV photons from of the emissions from the source. The half-value required for attenuation or complete absorption tor, it is %rst necessary to determine the thicknesBefore choosing the most appropriate attenua- t F shieldingF requirements areabout times16 Pb) to avoid whole-body exposure (Fig. 1) Shielded hot cell and dose calibrator (50 mm The thickness and density of the attenuator The nature andenergy ofthesource 99m Tc,whose HVL 18 F-FDG, as follows:

inlead is0.2 mm, 18 F,i.e. the s 27 Figure shields 3:Syringe Figure dispensing 2:Manual t t t tion when dispensing manually [17] (Fig. 2) doses (Fig. 3) of administration for used syringes the all %t to sizes in mm) (W≥5 shields Syringe inserted, inserted, providing more e(ective protec- is needle the which through aperture an with mm) (W≥20 caps and W,mm) 20-25 or mm (Pb≥30 pots Adequatedispensing administration room the to laboratory hot the from syringes ed Shielded trolley (Pb, 30 mm) to move shield-

Courtesy of Medisystem. Courtesy of Centre Leclerc.EANM t Lead mobile screen (≥30 mm) highly Figure 5a,b: Remote injection (Medisystem) re commended to reduce whole-body exposure when standing next to the patient (injection process, removal of cannula) or when operating the scanner from the gantry (Figs. 4–6)

Figure 4: Manual injection Courtesy of Centre Leclerc. Courtesy of Centre Figure 6: Operating camera from gantry

t Lead containers (≥10 mm) for waste and sharps Courtesy of Centre Leclerc. Courtesy of Centre

28 lowering %ngerdoses(Fig. 7). of means promising very a seem techniques of 18 preparation/injection the for reported studies of number The injection. and ration prepa- dose of process manual the replace to designed are systems aforementioned Intego sion of systems combining safe dispensing and infu- Automatic techniques. protective speci%c ministration process is then using performed and Althea Comecer Trasis UniDose). ad- The pensing process is of particular relevance (e.g. dis- the of automation manipulate, to heavy very are shields syringe and vials thick Since T, time;D, 20-40%attenuation >40%attenuation; Medium, distance;S,shielding;High, also be used for injections. In addition, shield- addition, In for injections. used be also good visibilities are now available; they should ing %nger doses [19]. Syringe shields providing a(ect- factor important most the is Shielding in out set Tableis techniques [18]. working 10 The relative importance of di(erent aspects of Chapter radiation 2:Practical inPET-CT protection Table ofprotection [18] factors impact 10:Likely Speed withwhichunshieldedmanipulationsare performed Position ofthe%ngershandholdingsyringe Shielding usedfor thevial shieldisusedWhether ornotasyringe Methodology F-FDG is limited. Nevertheless, automated Nevertheless, limited. is F-FDG TM 18 F-FDG are also available (e.g. MEDRAD and Lemer-Pax Posijet). Both of the of Both Posijet).Lemer-Pax and 29 handled directly. content. Unshielded syringes should never be their from emitted radiations ionising against environment the protecting [12], doses body contri ing whole-of optimisation the to butes MEDRAD Intego Figure 7:Automatic usingthe injection rtcinImpact Protection D + T S + T TM T S system Medium /HighMedium /HighMedium High High

Courtesy of Centre Leclerc. EANM Monitoring of occupational exposure Figure 8: Personal whole-body dosimeters The rapid increase in 18 F-FDG PET-CT studies raises the question of whether whole-body and extremity exposures are being maintained below the regulation limits. Close monitoring of sta( doses is required to con%rm that the highest level of protection is achieved.

All sta( working in a PET-CT facility must wear a personal dosimeter (OSL, TLD, %lm badge, electronic dosimeter) (Fig. 8). In addition, sta( preparing and injecting FDG doses should Leclerc Courtesy of Centre wear %nger dosimeters to demonstrate that Figure 9: Ring dosimeters equivalent doses do not exceed the annual limit of 500 mSv (Fig. 9). Regular exposure close to the limit is not in accordance with the fundamental ALARA principle. Occasion- ally, if an individual is found to be exposed at a consistently high level, close to the individual dose limit, so that the accumulated e(ective dose may be approaching an unacceptable level, then special attention should be given

to the optimisation of protection [15]. Leclerc Courtesy of Centre

Dosimeters should be worn routinely and at ing on which side of the hand the dosimeter appropriate positions [20]. An evaluation of element is worn (palm or back) [19]. working practices requires all users to wear their dosimeters every day and always in the Several studies investigating the doses re- same place. Ring dosimeters should be worn ceived by PET sta( are available [11, 12, 13, on the same %ngers and in the same orienta- 17]. Whole-body doses vary considerably be- tion with respect to the radiation source. The tween centres and depending on local work- most exposed parts of the hands are likely to ing practices. Roberts et al. [12] reported an be the tips of the index and middle %ngers. An estimated dose of 4.1 µSv per PET procedure, empirical multiplying factor may be applied to injection of FDG contributing the most to ra- doses recorded by ring dosimeters depend- diation exposure.

30 patients on the scanning bed results in 96% in results bed scanning the on patients of year. a Installation patients 2000 for 6 mSv to leading examination, per µSv 3 is period 3-year last the over MGP) XB, 2000 (DMC H dose, equivalent average the institute, our In Chapter radiation 2:Practical inPET-CT protection tion measures fundamental to the design of design the to fundamental measures tion This section considers short radiation protec- Introduction Meadows and PeterAngela Hogg DESIGN DEPARTMENT c) best patientcare andwelfare. compromising without limits regulation the low as reasonably achievable and well below medicine technologists can be maintained as nuclear and radiographers trained properly by received doses radiation The exposure. involving situations all in protection of level possible best the on achieve to e(ort going collective a requiring process iterative an is Optimisation exposure. sta( minimise to bust radiation safety programme are required ro- a and devices care.shielding Appropriate with manipulated be must tracers PET ment. day challenge in any nuclear medicine depart- Dealing with sta( exposure remains an every- procedures. and radiopharmacy including nuclear medicine cameras was 2.15±0.52 mSv, three and PET-CT through weekly rotating radiographers ten byreceived dose e(ective mean the 2009, In exposure. whole-body of p (10), measured with electronic dosimeters electronic with measured (10), 131 I-therapy 31 as with most modern PET-CT systems. PET-CT as withmostmodern requirement,x-radiationa of is use the when when using 511-keV radionuclides, particularly considerations layout and constructional important highlight will section section. This included suggested reading at the end of the reading,backgroundhave we important this in engage to wish you Should department. PET-CT a for requirements building detailed outline to or departments medicine nuclear camera gamma conventional of design the review to not is intention Our public. the of members and patients sta(, to dose diation ra- minimise to as so department PET-CT a ministration areas will also require designation and+ooring etc. (Fig. 10). Dispensing andad- surfaces administrationarea and dispensing notablein radioactivebewill these spillsand of spread the minimise to seek thatfeatures therewill berequirementa to include design medicine, nuclear camera gamma with As measurespractical needconsideration. of number a this, achieve to order In posed. ex- are public the and patients sta(, which to dose radiation of amount the minimising of importance the understand will employer regardless of geography, is that a responsible cility. However, what is of central signi%cance, centres and PET-CT the layout fa- of a PET-CT of construction the on advise will guidance ance relating to radiation protection, and the havewill guid- ownlaw and its country Each Regulation andguidance

EANM as either ‘controlled’ or ‘supervised’ in accor- if modernising an existing nuclear medicine dance with dose limits and guidance. Radiation imaging facility for this purpose. warning signs are also required as appropriate. Problems and solutions for 511-keV Figure 10: Smooth, easy to clean surfaces are photons appropriate in view of the risk of a radioactive spill. The fact that 511-keV photons are more pen- Radiation warning signs denote potential hazards etrating than the low- or medium-energy photons commonly associated with 99m Tc and 111 In has led to the imposition of additional department design requirements on PET-CT centres. Both distance and shielding must be used to good e(ect for the purpose of radiation dose reduction for sta( and the public.

In order to address this issue, as a %rst step the administration/uptake rooms can be made larger, particularly in length, so that people © A. Meadows Alliance Medical. passing by the doorway are less likely to be Department design must also extend to the subject to exposure from a radioactive patient. ‘designation’ of sinks, toilets and drains for the Figure 11 demonstrates a good example of disposal of limited quantities of radionuclides. this. Here the uptake couch is several metres It is essential that there are su#cient designat- from the entrance to the room, thereby mak- ed ‘hot toilets’ as a signi%cant proportion of the ing maximum use of distance to minimise dose administered patient injection (approximately to sta( and members of the public in nearby 30%) will be excreted by the patient as they corridors. Single occupancy uptake rooms empty their bladder prior to their PET-CT scan. are recommended as opposed to shared bay facilities since they will help to minimise the Generally speaking, the traditional gamma additional radiation dose to sta( and other pa- camera nuclear medicine department will tients and will also allow for ‘quiet time’ during not a(ord any radiation shielding within its the uptake period. Furthermore, optimising walls, and personnel will not have been re- the number of rooms will assist in optimising quired to stand behind secondary radiation scanner usage and ensuring e#cient through- shields during patient imaging procedures. It put. A busy city hospital could warrant up to is essential that such practical considerations four uptake rooms, although this judgment are taken into account for PET-CT, particularly is wholly dependent on potential workload.

32 © A. Meadows Alliance Medical. Chapter radiation 2:Practical inPET-CT protection typical typical dose rates from a patient in a particular ment, assessment must be made to measure layoutchosen the on pending depart- the of De- essential. are sta( protect to walls room scan the within materials shielding hour), 1 to (up complete to time of amount ni%cant sig- a take can scan PET the that Knowing unit. orpaediatric to amaternity next located not is it that ensure to prudent is it sidering location of a facility on a hospital site, con- when Furthermore, restricted. is access public where land of area enclosed an is 11 sible. For instance, beyond the rear wall in Fig. pos- where used be can again distance then ditional shielding is used in wall construction ad- unless and consideration require rooms uptake the of aspects external the Equally Figure 11: Typical administration/uptake room 33 trol and the scan room, demonstrated in Fig.in demonstrated room, scan the and trol con- the between wall the within contained high-energy photons are to be used. They are when construction wall in shielding ditional struction. Typically lead bricks are used as ad- con- wall in used be can that brick lead mm 15-a of example an demonstrates 12 Figure design.tion ofdepartment be required in the work-up and implementa- will professionals experienced of team a and mended shielding will depend on the facility, recom- and design Department rooms. the to provide additionalshielding Figure 12: A 15-mm lead brick; such bricks help demonstrated inthisimage. the glass window of the patient viewing area 13. In addition, lead salts are contained within area based on the dimensions and layout of layout and dimensions the on based area

© A. Meadows Alliance Medical. EANM Figure 13: Control room where shielding with X-ray considerations lead salt glass and lead bricks is used in the Fortunately, in terms of cost limitation, the adjoining wall to the scan room inclusion of lead within scan room walls and doors has a dual bene%t in shielding from x- rays produced by the CT scanner. The path of the x-rays must be considered and dose rates measured around the scan room and adjacent rooms during acceptance testing phases. A typical CT scatter plot image seen in Fig. 15 provides a visual representation of x-ray distribution during the CT element of the scan. This helps to demonstrate the importance of considering the distribution of shielding in

© A. Meadows Alliance Medical. room design and illustrates that, for example, An additional consideration is the imaging and shielding must be of signi%cant thickness be- uptake room doors: again, increased lead shield- tween the control and the scan room. ing is recommended, as illustrated in Fig. 14. Figure 15: Typical CT ‘scatter plot’ depicting the Figure 14: Side view of the scan room door. path of x-rays during exposure Lead lining approximately 4 mm wide can be seen and provides additional shielding © General Electric Company. All rights reserved. © General Electric Company. © A. Meadows Alliance Medical.

4 mm lead 34 © A. Meadows Alliance Medical. sign during CT x-ray exposuresign CT during Figure 16: External +ashing radiation warning Chapter radiation 2:Practical inPET-CT protection warning measurewarning access. to restrict occurring (Fig. 17) and provides an additional is exposure x-ray an that clear it makes also door room scan the over across’a ‘pulltape departments some In underway. is exposure warning sign (Fig. 16) to indicate that an x-ray ‘+ashing’ a include also scanner CT the by The additional design requirements imposed 35 restricted accessrestricted ensure to measures Additional 17: Figure conforms to allrelevant legislation andlaw. space and layout available and that the design the to relation in distributed appropriately is shielding that ensuring phase, design the at work-up the support protection radiation in experienced professionals of team a that the outset. Furthermore, it is strongly advised additional shielding and layout of the facility at givenbe resulttomust careful consideration a as and energies higher much with dealing protectionguidance. However, weare PET in radiation to adhere to measures as shielding and distance time, use we ensure we that in imaging, medicine nuclear conventional in those to similar are protection radiation of principles the fundamentally summary, In

References 13. Guillet B, Quentin P, Waultier S, Bourrelly M, Pisano P, 1. Townsend DW. Positron emission tomography/compu- Mundler O. Technologist radiation exposure in routine clinical ted tomography. Semin Nucl Med 2008;38:152-66. practice with 18F-FDG. J Nucl Med Technol 2005;33:175-9.

2. International Commission of Radiation Protection. Ra- 14. Council Directive 96-29 Euratom of 13 May 1996 laying diation dose to patients from radiopharmaceuticals. ICRP down the basic safety standards for the protection of the Publication 80. London: Pergamon Press; 1997. health of workers and the general public against the dan- gers arising from ionizing radiation 3. Lassmann M, Biassoni L, Monsieurs M, Franzius C. The new EANM pediatric dosage card: additional notes with re- 15. ICRP (International Commission on Radiological Protec- spect to F-18. Eur J Nucl Med Mol Imaging 2008;35:1666–8. tion). 1990 Recommendations of the International Commis- sion on Radiological Protection. ICRP Publication 60. Oxford: 4. Watson CC, Casey ME, Bendriem B, Carney JP, Town- Pergamon Press; 1991. send DW, Eberl S, et al. Optimizing injected dose in clinical PET by accurately modeling the counting-rate response 16. Delacroix D, Guerre JP, Leblanc P. Radionucléides et functions speci%c to individual patient scans. J Nucl Med radioprotection. Les Ulis: EDP Sciences; 2006. 2005;46:1825-34. 17. Prévot S, Touzery C, Houot L, et al. Optimization of tech- 5. Karp JS, Surti S, Daube-Witherspoon ME, Muehllehner nologists’ hands exposure: impact of vial shielding when G. Bene%t of time-of-+ight in PET: experimental and clinical preparing 18 F-FDG doses. Eur J Nucl Med Mol Imaging results. J Nucl Med 2008;49:462-70. 2008;35:Suppl 2:T13.

6. European guidelines on quality criteria for computed 18. Martin CJ, Whitby M. Applications of ALARP to extremity tomography (EUR 16262 EN, May 1999). doses for hospital workers. J Radiol Prot 2003;23:405-21.

7. Arch ME, Frush DP. Pediatric body MDCT: a 5-year follow- 19. ICRP (International Commission on Radiological Protec- up survey of scanning parameters used by pediatric radio- tion). Radiation dose to patients from radiopharmaceuticals. logists. Am J Roentgenol 2008;191:611-7. ICRP Publication 106. New York: Elsevier; 2009.

8. McCollough CH, Zink FE, Ko+er JM, Matsumoto JS, Tho- 20. Donadille L, Carinou E, Ginjaume M, Jankowski J, Rimpler mas KB. Dose optimization in CT: creation, implementation A, Sans Merce M, et al. An overview of the use of extremity and clinical acceptance of size-based technique charts. dosimeters in some European countries for medical appli- RSNA 2002 Scienti%c Program, Supplement to Radiology cations. Radiat Prot Dosimetry 2008;131:62-6. 2002;225:591. Suggested reading 9. Alessio AM, Kinahan PE, Manchanda V, Ghioni V, Aldape Christian P, Waterstram-Rich K. Nuclear medicine and PET/ L, Parisi MT. Weight-based, low-dose pediatric whole-body CT – techniques and technology. 6th ed. St. Louis: Mosby PET/CT protocols. J Nucl Med 2009;50:1570–8. Elsevier; 2007.

10. Fahey FH, Palmer MR, Strauss K, Zimmerman RE, Badawi European Guidelines on Quality Criteria for Computed To- R, Treves ST. Dosimetry and adequacy of CT-based attenua- mography. EUR 16262, EU 1998. tion correction for pediatric PET. Radiology 2007;243:96-104. Martin C, Sutton D. Practical radiation protection in health- 11. Benatar NA, Cronin BF, O’Doherty MJ. Radiation doses care, Oxford: Oxford University Press; 2002. rates from patients undergoing PET – implications for tech- Saunders SE. Computed tomography: physical principles, nologists and waiting areas. Eur J Nucl Med 2000;27:583-9. clinical applications, and quality control . 3rd ed. Philadel- 12. Roberts FO, Gunawardana DH, Pathmaraj K, Wallace phia: W.B. Saunders; 2008. A, U PL, Mi T, et al. Radiation dose to PET technologists Seerum E. Rad Tech’s guide to radiation protection. Malden, and strategies to lower occupational exposure. J Nucl Med Mass.: Blackwell Science; 2001. Technol 2005;33;44-7. Statkiewicz Sherer MA, Visconti PJ, Ritenour ER. Radiation protection in medial radiography. 5th ed. St. Louis: Mosby; 2006

36 cose(FDG) theismost commonly andwidely positron-emittingtracers. tive in vivo measurements of 3D distributions of BoellaardRonald as will be explained in more detail later. based on ‘standardised uptake values’ (SUVs) [7], prognosticaindicator as PET of use the and the assessment of tumour response on therapy in which quanti%cation of PET is important are titative manner more extensively. Applications modality it is likely that it will be used in a qua Nevertheless,quantitativeais PETas imaging tumour staging and patient management [5, 6]. spection of whole-body FDG images is used for information.Typically, oncology,in in-visual su#cient provides images PET of spection applications,thesein-visualof For somethe (FWHM) for clinical PET scanners]. days up to ~2.5 mm full-width at half-maximum sensitivitywithhighspatialresolution [nowa- of new drugs [3, 4]. PET imaging combines high application or for the evaluation of the e#cacy assessment of therapeutic responses as a clinical expression.Furthermore, usedforbecan PET and a#nity, drug delivery and uptake and gene oxygenconsumption, neuroreceptor density may be derived, such as blood +ow, glucose and physiologicalpharmacokineticor parameters nancies [1, 2]. By using di(erent tracers, various sumption and it is mainly used to detect malig- basicallyprovidesmeasure glucosea ofcon- used PET tracer in oncological applications. FDG protocol optimisation ofPET principles imaging instrumentationand Chapter 3:PET calimaging technique which allows quantita- Positron emission tomography (PET) is a medi- Introduction 18 F-+uorodeoxyglu- n- 37 PET imaging for studies.PET multi-centre the a(ectingfactors SUV and optimisation of on focus will section second Thepresented. be %rst will instrumentation PET and PET of this paper some background on the principles In studies. PET of evaluation during account should be considered carefully and taken into quanti%cation regarding PET of Limitations quanti%cation. a(ect may parameters these extent what to realise to importance utmost struction and data analysis. It is therefore of the ods used during PET acquisition, image recon- pends on parameters/settings and the meth- de- however,studies, PET of Quanti%cation comfort (e.g.comfort claustrophobia). improvepatient and/or access patient easier forallow may option latter apart. Themoved be to systems two the allowing covers, rate sepa- two in or coverlarge one within either part PET the of front in placed is scanner CT tial, but integrated, system design in which the Healthcare. All PET-CT systems have a sequen- Medical, Toshiba Medical Corporation and GE Healthcare, Siemens Medical Solutions, Hitachi Philips systems: PET-CT o(ering vendors %ve are there present At [8]. Townsendand lawi Maw- by Imaging Molecular and Medicine published in the European Journal of Nuclear overview on technologyPET-CT was recently excellent An systems. PET-CT multimodality combined are systems PET most Nowadays andPET/MRI PET-CT imaging andinstrumentationPET

EANM At present, clinical (prototype) PET-MRI sys- Principles of PET tems are being built with di(erent designs, as PET is a molecular imaging technique which follows: (a) a PET insert placed within the MRI measures the distribution of a radioactive scanner, thereby allowing for truly simultane- tracer in vivo [12]. Upon administration of ous PET and MRI acquisitions, but restricting very small amounts (pico- or nanomoles) of a the system’s application to brain imaging; (b) radiotracer to the patient it distributes among a design in which MRI and PET are placed side and within the organs. The radioactive atom by side, i.e. a similar arrangement to that used of the radiotracer emits positrons. The emit- for PET-CT systems. In the latter case, acquisi- ted positron combines with an electron after tions will be nearly but not exactly simultane- travelling a distance up to several millimetres ous, but these systems allow for whole-body in tissue. The positron and electron are then acquisitions. Major challenges with PET-MRI converted into two photons, each having an acquisitions are the development of PET de- energy of 511 keV, which are emitted in nearly tectors that are insensitive to the magnetic opposite directions. PET image acquisition is %eld of the MRI scanner and use of MRI data based on the simultaneous (coincidence) de- for attenuation (and scatter) correction of the tection of these two photons. A PET scanner PET data. Much progress has been reported consists of many photon detectors surround- in addressing these issues [9]. ing the patient. During a PET scan millions of coincidence detections are collected, provid- The remainder of this paper will, however, focus ing information about the distribution of the on PET(-CT) imaging as PET-CT is widely avail- radiotracer in tissue. Figure 1 demonstrates able and used in a routine clinical setting [10, 11]. the principles of PET imaging.

Figure 1: Principles of PET imaging Courtesy of Hedy Folkersma, Courtesy of Hedy Folkersma, VU University Medical Centre, Netherlands. The Amsterdam,

38 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

PET camera will notice a random coincidence detection. It may be clear that these random coincidences result in image distortions (appear- ing as the addition of a smooth background). Finally, multiple detections can occur when three or more photons are detected at the same time. These multiples are usually discarded.

Quanti%cation of PET studies requires that the contributions of scattered and random coincidences are accounted for. Moreover, due to attenuation (=scatter and absorption) of photons © R. Boellaard. EANM Figure 2: Illustrations of true (top, left), random in the patient, a large fraction of the emitted (top, right), scatter (bottom, left) and multiple photons is not detected. Fortunately, in PET, at- (bottom, right) coincidences tenuation does not depend on the location of the positron emission along the line of response, Unfortunately, not all coincidences contribute to i.e. the line connecting the detectors where a the signal, i.e. the ‘true’ 3D distribution of the trac- coincidence is measured. Consequently, by er. Background noise is added to the signal due acquiring transmission and/or CT scans, the ef- to photons that are scattered before detection fects of attenuation can be corrected for exactly. or by coincidence detection of two uncorrelat- In practice, however, attenuation correction is ed photons, i.e. so-called random coincidences. somewhat hampered by patient motion. Ran- Figure 2 illustrates the di(erences between true, dom, scatter and attenuation correction meth- random, scatter and multiple coincidences. True ods will be discussed later in more detail. coincidences arise from the simultaneous (coincident) detection of two annihilation photons Acquisition and image reconstruction: generated by one positron emission. Ideally, only 2D versus 3D true counts are detected. A large fraction of the Although PET is a 3D imaging method, in the emitted photons (up to 50%) is scattered before past a lot of PET and PET-CT scanners were leaving the patient. When one of the photons has equipped with septa, i.e. lead or tungsten an- been scattered, it will result in a dislocation of the nular shields positioned within the %eld of view ‘true’ coincidence detection. Moreover, when two (FOV). These septa served to shield the detectors photons from two di(erent positron emissions from photons emitted or scattered outside the are accidentally (randomly) detected simulta- transverse or transaxial plane (Fig. 3). The main neously (while the others are undetected), the purpose of using these septa (2D mode) was

39 Figure 3: Illustration of 2D (left) and 3D (right) acquisitions. Red crosses indicate photons that are shielded from the detectors by the septa. Indices above and below each ‘detector’ indicate corresponding types of coincidences Courtesy of Mark Lubberink, VU University Medical Centre, Netherlands. The Amsterdam,

to reduce the contribution of random coinci- tions have a higher detection probability, result- dences, scattered photons and photons coming in increased sensitivity but also in increased ing from activity outside the FOV at the cost of random and scatter contributions. Fortunately, reduced sensitivity compared with 3D (no septa) most of these scanners are also equipped with acquisitions. It was generally considered that the new and fast detectors using new scintillation reduced contribution of random and scattered crystals (Table 1). These new crystal materials photons in the case of 2D acquisitions improved result in better count rate performance as they the quantitative accuracy of PET studies, al- show a faster scintillation rise and decay time. though this is nowadays a matter of debate [13]. Consequently, a shorter coincidence time win- Most modern PET and PET-CT scanners are no dow can be applied, resulting in a reduction of longer equipped with septa, and acquisition in random (and scatter) coincidences. Use of these 3D mode (without septa) is the only option. 3D new crystals therefore partly compensates for acquisition includes lines of response (LORs) that the increase in random and scattered photons are located in oblique planes (Fig. 3). 3D acquisi - with 3D acquisitions.

Table 1: Characteristics of common scintillation crystals Property NaI BGO LSO GSO Density (g/ml) 3.67 7.13 7.4 6.7 E(ective Z 51 74 66 61 Decay time (ns) 230 300 35-45 30-60 Photons/MeV 38000 8200 28000 10000 BGO, bismuth germinate; GSO, gadolinium oxyorthosilicate; LSO, lutetium oxyorthosilicate; NaI, sodium iodide

40 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

Random correction Attenuation correction As indicated above, random coincidences Attenuation of photons in the patient causes arise from the ‘simultaneous’ detection of loss of coincidences. The number of measured two uncorrelated photons, i.e. photons com- coincidences depends on the patient’s ‘radio- ing from two di(erent positron emissions. As logical’ thickness. The e(ects of attenuation these randoms are uncorrelated, their mu- can be easily compensated for by acquiring tual directions are also uncorrelated and an transmission or CT scans. In short, during almost uniform background is added to the transmission scans the radiological thickness measured 3D activity distribution. of the patient is measured for each possible LOR. The transmission through the patient is Nowadays most PET-CT systems use sophisti- obtained by taking the ratio of the measured cated random correction methods based on transmission scan counts to those obtained EANM using either block singles or a delayed coinci- during a blank scan, i.e. a transmission scan dence time window method [12]. A detailed acquired without the patient in the FOV. This explanation of these methods is beyond the ratio (transmission) is a direct measure of the scope of this chapter. In short, the delayed attenuation loss per LOR. Correction can be coincidence time window method is based performed by multiplying a ‘trues’ sinogram on counting coincidences using a time-shifted by this (inverse) ratio, although more sophis- coincidence window. As the time window is ticated implementations within the recon- shifted, measured coincidences come from struction algorithm are usually applied [14]. photons of uncorrelated positron emissions Occasionally it may be necessary to compen- by de%nition. The ‘measured’ randoms using sate for emission spillover, i.e. emission counts this technique provide an accurate estimate of measured during transmission scanning, and/ the randoms distribution and contribution. A or to enhance the quality of the attenuation randoms-corrected sinogram can then be ob- correction using dedicated segmentation al- tained by subtracting this randoms sinogram gorithms – an approach usually referred to from the measured total coincidence sono- as transmission image segmentation or seg- gram, resulting in a ‘trues’ sinogram. In most mented attenuation correction [15]. cases this correction is performed online during acquisition, although randoms correction In PET-CT systems the CT data are used to can be applied in a more sophisticated and derive the attenuation correction (CT-AC). CT accurate way during image reconstruction. scans do not su(er from a poor signal-to-noise ratio (SNR) compared with ordinary transmission scans and emission spillover. A CT scan is made by rotating an x-ray tube around the

41 patient. The energy of the photons generated lected over various respiratory cycles, while by the x-ray tube is much lower than 511 keV. A the CT-AC provides a series of images each CT image therefore represents the distribution acquired during only a short phase of the re- of attenuation coe#cients for lower photon spiratory cycle. CT-AC images therefore do not energies (~70 to 90 keV). The coe#cients are match the emission data and some interface not directly applicable for 511 keV and the CT artefacts may appear. A common strategy to image needs to be converted into a “511-keV reduce (but not to avoid) these artefacts is to attenuation coe#cient image”. Moreover, CT instruct the patient to breathe shallowly. scans may su(er from beam hardening artefacts as the x-ray beam consists of a spectrum Finally, CT-AC may provide incorrect attenua- of photon energies and the attenuation de- tion correction in a number of special cases. pends on the photon energy. A single scaling For example, special attention is needed in the factor may not be used to ‘convert’ the CT scan case of metal implants. While these implants into a “511-keV attenuation coe#cient image”. seem relatively opaque on 511-keV transmis- Normally this conversion is performed by ap- sion scans, they give rise to severe artefacts on plying conversion functions per tissue class (e.g. CT-AC images. When an implant is overlooked, bilinear scaling functions) to rescale Houns%eld there will be a false impression of an increased units (HU) to “511-keV attenuation coe#cients”. signal on the PET scan at the location of the Next, this rescaled image is used to derive the implant. Similarly, use of contrast agents pro- attenuation correction factors per LOR. duces a high-intensity signal on the CT scan and results in an incorrect attenuation cor- The accuracy of both transmission scan- and CT rection, although it seems that with current scan-based attenuation correction may be af- CT-AC processing algorithms the impact of fected by patient motion. Clearly, any displace- contrast agents on the accuracy of CT-AC is ment of the patient between the transmission moderate. Finally, in some PET-CT scanners the or CT scan and the emission scan causes a spa- FOV of the CT is smaller than that of the PET tial mismatch between the two data sets and scanner. When the patient is only partially vis- results in an incorrect attenuation correction. ible on the CT scan, e.g. sometimes the arms Patient motion should therefore be restricted of the patient are truncated, it is not possible as much as possible. Clearly, respiratory mo- to calculate the attenuation for those LORs tion cannot be avoided. E(ects of respiratory passing through the arms fully correctly and motion can be quite pronounced in the case attenuation correction could be wrong. A re- of CT-based AC. A CT scan provides almost a view on the use of PET-CT and the limitations ‘snapshot’ of the patient as a (spiral) CT scan of CT-AC has been provided by Von Schulthess is acquired very rapidly. Emission data are col- et al. [16].

42 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

Scatter correction combination with the transmission scan data The above information on random and atten- to calculate a scatter sinogram. The method uation corrections might seem complicated, assumes that scatter is mainly caused by and it is true that many factors can in+uence single scatter events of one of the photons their accuracy. Yet, development of accurate and it tries to simulate scatter using the fact scatter correction algorithms is still one of that almost all scattered events are caused by the most challenging topics in PET physics, Compton scattering. Compton scattering has especially in the case of 3D scans (no septa). been characterised well and is used within As most modern systems are 3D only, this sec- the simulation. A detailed description of this tion will summarise scatter correction issues method can be found in [17]. As the method for 3D scans only. is based on the physical principles of photon scattering and takes distribution of activity EANM Scattered coincidences occur when one or and attenuation into account, it seems to be both photons are scattered and thereby de- an accurate scatter correction method, and is +ected from their original direction (Fig. 2). The nowadays routinely available. main cause of scattering events for 511-keV photons is Compton scattering, where the Image reconstruction methods photon interacts with (‘hits’) an electron. Af- PET measures coincidences, which are usu- ter scattering, the photon has a lower energy ally stored in sinograms. A sinogram contains (depending on the angle of de+ection from its the projections over all angles of the activity original direction) and it obtains a di(erent di- distribution in the patient. The process of cal- rection. Scattering results in an almost random culating the 3D activity distribution in the pa- direction of the scattered photon (although tient from the measured sinograms including forwardly peaked) and adds a low-frequency correction for randoms, scatter, attenuation, background onto the image (thereby reduc- normalisation and dead time is called image ing contrast). reconstruction. Image reconstruction algorithms can be classi%ed into analytical and it- The most frequently applied scatter correc- erative methods and into 2D and 3D methods. tion method in PET is based on estimating the scatter distribution/contribution using a The most commonly used analytical image re- single scatter simulation method described construction method is %ltered back-projection by Watson et al. [17]. This method initially re- (FBP). This method is linear and quantitatively constructs a %rst estimate of the distribution robust. However, the method is sensitive to of the activity in the patient without scatter noise and reconstructed images may contain correction. Next, this initial estimate is used in severe streak artefacts. For these reasons, itera-

43 tive reconstruction algorithms have been developed such as ordered subset expectation maximisation (OSEM). Other iterative methods have been developed as well, such as least square [18] and RAMLA [19]. In general, during iterative reconstruction an image is generated by repeatedly (iteratively) estimating an image Courtesy of © R. Boellaard, the Department of Nuclear Medicine & PET Research, VU University Medical Centre, Netherlands. The Amsterdam, and its corresponding sinogram. Iterations are Figure 4: Examples of whole-body FDG PET continued until there is an optimal match be- images. Left, images reconstructed using tween the estimated and the measured sino- FBP without attenuation correction; middle, gram. The drawback of these methods is that FBP with attenuation correction; right, OSEM both quantitative accuracy and SNR depend on iterative reconstruction with attenuation the number of iterations [20]. Too few iterations correction result in quantitative inaccuracy (no convergence), while too many iterations amplify noise A second classi%cation of image reconstruc- to unacceptable levels. A trade-o( has to be tion methods is based on 2D or 3D reconstruc- found for each speci%c application. Moreover, tions. In both cases 3D volumetric information in the case of iterative reconstruction, conver- about the activity distribution in the patient is gence (i.e. the image reaching the ‘true’ activity obtained. 2D reconstruction methods recon- distribution) depends on the underlying source struct images plane-wise. When acquisitions distribution, and optimal reconstruction set- are performed in 3D, the 3D sinogram con- tings may therefore be di(erent for di(erent taining oblique planes is %rst converted into types of PET study (body part under investiga- a 2D sinogram containing axial planes only. tion or di(erent scan statistics). For oncological The latter process is called rebinning. Fourier whole-body studies, in comparison with FBP, rebinning (FORE) [23] is generally used for this OSEM reconstruction clearly provides images purpose and is presently the most accurate with better (visual) image quality (Fig. 4) and rebinning method implemented for most with almost equal quantitative accuracy [14, systems. After (Fourier) rebinning, a 2D sino- 21, 22]. However, in the case of dynamic PET gram is obtained and reconstructions can be studies consisting of many frames with short performed plane by plane (2D). Reconstruc- scan durations and thus poor statistics, iterative tion speed was the main rationale of using reconstruction may show biases (i.e. quantita- rebinning in combination with a 2D recon- tive inaccuracies). Consequently, FBP is often struction. The drawback of rebinning is that still the preferred reconstruction method for it reduces the resolution at increasing radial dynamic PET studies [20]. distance from the centre of the axial FOV. This

44 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

resolution-degrading e(ect of FORE depends the positron emission and taking its uncer- on and increases with the axial aperture of the tainty into account within the reconstruction scanner and sinogram rebinning is therefore method, ToF reduces image noise and seems not feasible for all scanners. to enhance contrast recovery. In other words, ToF is presently used to improve image quality. To avoid resolution loss due to rebinning, 3D reconstruction methods are nowadays applied A second new development, which has re- which use the full 3D sonogram or all LORs. As a cently also become available on clinical scan- 3D sinogram contains many more LORs, it can ners, is the use of recovery correction during be easily understood that fully 3D reconstruc- image reconstruction [25]. This image recon- tions are computationally more demanding. struction method uses the (measured) spatial- Modern PET-CT systems are therefore equipped ly variant point spread function, either image EANM with dedicated computer clusters to reconstruct or sinogram based, during the reconstruction the image within a reasonable time and fully 3D process to reduce partial volume e(ects and reconstruction has become the standard. thereby enhance the spatial resolution of the reconstructed images. First clinical evaluations Recent technologies: time of +ight and resolu- using these methods have been published tion recovery [26]. Further evaluation of the impacts on In the past few years, PET-CT systems with quanti%cation and on image quality is war- time of +ight (ToF) capabilities have become ranted, but %rst results are promising. commercially available [24]. ToF is based on the di(erence in arrival or detection time of Optimisation of PET imaging for multi- both 511-keV annihilation photons when the centre study quanti#cation annihilation took place at o(-axis locations. As pointed out above, accurate corrections The di(erence in detection time contains in- to account for many factors at a technical or formation about the position of the positron data collection level, such as random coinci- emission along the LOR. ToF requires a high dences, scattered photons, attenuation e(ects timing accuracy and fast detectors with high and dead time, have been developed and are sensitivity and fast electronics, which have being applied in most modern PET-CT sys- only recently become available. At present, tems. PET is therefore essentially a quantitative assessment of the exact location (within a medical imaging technique that can measure couple of millimetres) with ToF is not feasible. the distribution and uptake of a radiotracer The current ToF technology provides a posi- quantitatively in vivo. Moreover, PET provides tional accuracy within about 10 cm FWHM. a quantitative measure of the underlying biol- However, by using the estimated position of ogy, such as metabolism, receptor density or

45 occupancy, transporter activity or information Factors a(ecting SUV quanti%cation on signaling pathways, depending on the ra- Although PET is a quantitative imaging tech- diotracer being used. FDG is presently widely nique, there are still many factors that a(ect used in the clinic and provides a quantitative quanti%cation of FDG PET-CT studies using index of glucose metabolism. SUVs. These factors have been described in [30] and are summarised below. The aver- High rates of glucose metabolism are associ- age or estimated magnitude of the impact ated with malignancy and FDG PET-CT stud- of these factors on SUV variability is indicated ies are therefore used for staging, prognosis in parentheses: and response monitoring purposes (using changes in glucose metabolism as a mea- Biological factors sure of tumour response). There is more and t Uptake period (15%) more evidence that quantitative measures of t Patient motion and breathing (30%) FDG uptake or its change can be used as a t Blood glucose levels (15%) prognostic factor, for response assessment or as a surrogate endpoint for therapy out- Technical factors come evaluations [27, 28, 29]. Widespread t Relative calibration between PET scanner use of quanti%cation of FDG uptake has been and dose calibrator (10%) hampered, however, by the vast variability in t Residual activity in syringe (5%) methodology applied to derive quantitative t Incorrect synchronisation of clocks (10%) measures of FDG uptake, such as the stan- t Injection vs calibration time (10%) dardised uptake value (SUV). The outcome t Quality of administration (50%) of SUV depends on many factors, as recently pointed out in the supplement issue of the Physics/data analysis-related factors: Journal of Nuclear Medicine [30]. As a con- t Scan acquisition parameters (15%) sequence, conclusions drawn based on SUV t Image reconstruction parameters (30%) data obtained in one centre are not valid for t Use of contrast agents (15%) studies performed elsewhere. In this section, t Region of interest (ROI) or volume of inter- the factors that a(ect SUV quanti%cation will est (VOI) method (50%) be brie+y discussed, followed by an explanation on how to optimise FDG PET-CT scanning Di(erent types of PET-CT system from di(er- procedures for use in multi-centre studies, i.e. ent vendors are being used today. These scan- to make SUV data exchangeable among in- ners have di(erent hardware con%gurations stitutes [31]. (detectors, scintillator material, electronics etc.) and use di(erent software and algo-

46 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

rithms for image reconstruction, corrections Patient preparation procedures and data analysis. These di(erences will not Patient preparation procedures describe all be overcome and di(erent scanners show measures to be taken into account prior to di(erent quantitative performances. Yet, vari- FDG administration and the PET-CT study. ability of SUVs in multi-centre studies can be Adequate patient preparation is needed to substantially minimised by taking a number maximise uptake in tumours and minimise up- of precautions and by employing procedures take in healthy tissues, thereby optimising PET that minimise the variability of SUV caused by study image quality for both diagnosis and the above-mentioned factors. Optimisation of quanti%cation. Below the two most important image quality for exchangeability and com- issues that a(ect the clinical procedure from a parability of SUV measures is thus based on practical point of view are discussed. principles that aim at reducing SUV variability EANM across sites [6, 32, 33, 34]. Standardisation of FDG uptake varies over time. Therefore the PET procedures addresses: time interval applied between FDG administration and the start of the PET study must t Patient preparation procedures be matched as closely as possible between scans performed at various sites. Generally an t FDG administration procedures interval of 60 min with a tolerance of +/- 5 min is considered acceptable. When PET studies t PET study statistics, image quality and SNR are performed for response monitoring purposes, an appropriate interval between the t ‘Clinical image’ resolution/contrast recovery end of the therapy cycle and the PET study needs to be considered as FDG uptake may t Data analysis procedures and SUV normali- vary strongly shortly after (chemo-)therapy sation [6]. The optimal interval is study speci%c and requires further investigations [3]. t Speci%c multi-centre quality control measures As both glucose and FDG are actively trans- ported into cells, glucose levels in blood af- These speci%c topics will be discussed brie+y fect the uptake of FDG and thereby the SUV. below, with a focus on the practical conse- High blood glucose levels will result in lower quences. It is recommended that interested uptake of FDG and thereby lower SUVs. When readers consult several other papers in which not properly taken into account, a high blood these items are outlined in more detail [2, 21, glucose level may erroneously result in (in- 35, 36, 37, 38, 39, 40]. correct) lower SUV data and will also hamper

47 visual interpretation of the PET images. It is %ed; (b) injection or administration time and therefore important that blood glucose levels (c) start of the PET-CT acquisition. The di(er- are within a normal range before FDG admin- ence between injection time and start of the istration. Usually normal blood glucose levels PET-CT acquisition provides the uptake period, (<7 mmol/L) can be reached by 4–6 hours’ which should be as close as possible to 60 min. fasting prior to the PET examination. Blood The time di(erence between dose calibration glucose levels should be checked before ad- time and PET-CT acquisition time is needed ministration of FDG. If blood glucose values to derive the decay-corrected FDG dose at are elevated (>7mmol/L), the PET-CT study the start of the PET-CT study. Alternatively, the should preferably be rescheduled [31], if clini- FDG dose at injection time may be entered in cally feasible. the PET-CT system during setup of the patient acquisition. In the latter case, decay correction FDG administration must be applied for the interval between FDG The net administered FDG dose needs to be dose calibration (or assay) time and injection known exactly. Paravenous injection should time, assuming that the scanner software then be avoided and residual activity in the syringe accounts for decay between injection and or administration system should be minimal scan start time (which should be checked). (<3%) or must be measured so that it can be accounted for. It is recommended to implement Image quality proper administration procedures that ensure The quality of a PET study depends on many that the net administered dose is known. Dis- technical factors, which have been addressed crepancies in assumed versus true net admin- in the %rst section of this chapter. In clinical istered dose will result in incorrect SUV data. practice, di(erences in image quality may oc- Calculation of net administered dose should cur due to di(erences in scanner sensitivities, also include appropriate corrections for decay relative bed overlap between subsequent and requires accurate synchronisation of clocks bed positions and patient weight. These fac- throughout the department. Decay correction tors could therefore also increase variability should be applied between the FDG dose cali- in SUV between scanner, institutes and pa- bration time and the start time of the PET study. tients. Moreover, poor scan statistics result in an upward bias of SUV [35]. Di(erences in scan It should be noted that three time points are statistics amongst centres and subjects may essential for correct SUV assessments: (a) FDG be minimised by prescribing FDG dosage as a dose calibration time or dose assay time, i.e. function of patient weight, relative bed over- the time at which the amount of FDG (MBq) lap of subsequent bed positions and emission that is to be administered to a patient is speci- scan acquisition mode (2D vs 3D) and acquisi-

48 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

tion duration (per bed position). The EANM e(ects are not widely available or validated/ guidelines for oncological FDG PET-CT studies approved. Contrast recovery can be achieved provide recommendations for FDG dose [31]. by strict prescription of reconstruction settings Although these recommendations are an im- per type of scanner and should be determined provement over a +at dosing procedure, i.e. all using dedicated QC phantom experiments. subjects receive the same FDG dose regardless of type of scanner and patient weight, further Data analysis procedures and SUV normalisation optimisation of FDG dose as a function of the The %rst step in deriving SUV is the assess- above-mentioned parameters is needed and ment of tracer uptake by placing an ROI over may be scanner dependent. or in the tumour as seen in the PET-CT images. Various ROI strategies can be applied, ‘Clinical’ image resolution or contrast recovery such as manually de%ning 2D and 3D ROIs or EANM The resolution and/or contrast recovery seen semi-automatic ROI generation, %xed size ROIs in clinical practice is determined to a large and use of the maximum intensity voxel. All extent by the reconstruction settings ap- ROI strategies have speci%c disadvantages and plied. Iterative reconstruction algorithms are bene%ts regarding ease of use, accuracy and being used mostly for reconstruction of FDG precision. Use of the maximum voxel value whole-body PET studies. Various parameters might be attractive as it is less dependent on of these algorithms, such as number of itera- the performance of manual or semi-automatic tions and subsets, relaxation factors, voxel size ROI procedures, but it may su(er from upward and post-reconstruction image %lter settings, bias in the case of increased noise levels [35]. determine the ‘clinical’ contrast recovery seen in Clearly, the uptake (SUV) derived from the PET practice [41, 42]. Moreover, a su#cient number study depends on the ROI methodology and of iterations (or its product with the number a method should be used consistently across of subsets) is needed to ensure su#cient con- all scans and institutes in a multi-centre study. vergence of image reconstruction. Insu#cient convergence results in a lower contrast recov- In most cases SUV is normalised by body weight ery (and thus lower SUV) and makes lesion SUV (SUV-BW), as indicated in Eq. 1. Other normalisa- more dependent on that of its surrounding. tions are also used, such as lean body mass and Di(erences in contrast recovery are probably body surface area [37]. The most appropriate one of the main factors contributing to vari- normalisation factor is, however, still a matter of ability of SUV amongst centres [21]. Matching of debate. Therefore, it is recommended that pa- contrast recovery across centres and scanners tient height should be measured in addition to is therefore essential in multi-centre studies as patient weight in order to allow for application long as methods to correct for partial volume of all the various SUV normalisations. Moreover,

49 the SUV calculation may include a correction for Equation 1 plasma glucose level (Eq. 2). However, although in theory more accurate results may be obtained by correcting SUV for plasma glucose, the introduction of an additional correction factor into the SUV calculation may worsen reproducibility, especially when not properly measured [43].

Equation 2

In Eqs. 1 and 2 ACvoi represents the average All scanners are equipped with (semi-)auto- activity concentration within a VOI over the mated procedures for daily quality control of tumour, FDGdose is the net administered dose the PET-CT system. The test usually reports corrected for decay between dose calibration hardware failures or drifts resulting in unac- time and start time of the PET study and BW ceptable image quality loss. The procedures represents the measured body weight. In Eq. are scanner speci%c and provided by the 2 the plasma glucose level (Pglu) is normalised manufacturer. In the case of PET-CT scanners, by a population average value of 5.0 (round- all daily tests should be performed for both ed-o( value). the PET and the CT components of the scanner. Clearly, all tests should be passed without Speci%c multi-centre quality control measures errors before (any) clinical use. The quality control measures speci%c for multi- centre PET studies should focus on three items: The relative calibration of the PET-CT scanner (a) correct functioning of the PET or PET-CT against the dose calibrator used for measur- camera according to speci%cations; (b) accu- ing patient FDG dose provides information rate (within 10%) relative calibration of the PET about potential discrepancies in the calibra- or PET-CT scanner against the dose calibrator tion of PET-CT and that of the dose calibra- used for measuring patient FDG doses and (c) tor. Cross-calibration (and its veri%cation) is veri%cation of activity concentration or con- equally important as the calibrations of the trast recovery as a function of sphere size to individual devices themselves. This can be eas- assure resolution or contrast recovery match- ily understood from Eq. 1. In the SUV calcula- ing amongst centres in a multi-centre study. tion the FDG uptake measured with PET enters

50 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

the equation in the nominator, while the in- The absolute values of the contrast recovery jected dose measured using a dose calibrator coe#cients and their relative change with is used in the denominator. Consequently, any sphere size provide a good measure of the discrepancy in absolute calibration results in overall partial volume e(ects seen under incorrect SUVs. For correct SUV data, an ac- conditions which are clinically more relevant curate relative (cross-)calibration is therefore [21]. When di(erent systems provide similar even more important than the accuracy of (absolute) recovery coe#cients measured in the (separate) calibrations of the individual a standardised way, resolution (and partial vol- devices (dose calibrator and PET scanner). ume e(ects) is su#ciently matched to allow interchangeability of clinical SUV data across Di(erences in spatial image resolution or institutions/centres. contrast recovery between various PET-CT EANM systems in a multi-centre study make a large Summary contribution to inter-institute SUV variability PET is essentially a quantitative imaging mo- [21]. Prescriptions for acquisition and recon- dality. This chapter has %rst described some struction parameters may be de%ned for background on the principles of PET and PET each type of scanner to ful%l resolution and instrumentation, thereby providing the reader convergence criteria. However, these pre- with a fair understanding of the quantitative scriptions will become obsolete with ongo- nature of PET. As quanti%cation of PET studies ing development of new PET-CT scanners also depends on the methods and procedures and (reconstruction) software. Moreover, used during PET acquisition, image recon- resolution is generally measured using point struction and data analysis, it is of importance sources, which may provide an optimistic es- to understand the e(ect of such factors on the timation of ‘clinical’ resolution in the case of main clinical quantitative parameter, the so- iterative reconstruction methods. Therefore called standardised uptake value. The second contrast recovery coe#cients provide a more part of this chapter has therefore focussed on clinically relevant measure of resolution and explaining a few of the main factors a(ecting convergence. The activity concentration re- SUV and (corresponding) procedures for opti- covery coe#cient is the ratio between FDG misation of quantitative PET imaging. uptake in a sphere measured by the PET-CT system compared with the real FDG uptake. Acknowledgements Recovery coe#cients can be measured using Hedy Folkersma and Mark Lubberink are phantoms containing variously sized spheres. thanked for providing illustrations.

51 References Chapter 3

References 14. Lartizien C, Kinahan PE, Swensson R, Comtat C, Lin M, 1. Smith TAD. FDG uptake, tumour characteristics and re- Villemagne V, et al. Evaluating image reconstruction me- sponse to therapy: a review. Nucl Med Commun 1998;19:97- thods for tumor detection in 3-dimensional whole-body 105. PET oncology imaging. J Nucl Med 2003;44:276-90.

2. Fletcher JW, Djulbegovic B, Soares HP et al. Recommen- 15. van der Weerdt AP, Boellaard R, Knaapen P, Visser CA, dations on the use of F-18-FDG PET in oncology. J Nucl Med Lammertsma AA, Visser FC. Postinjection transmission 2008;49:480-508. scanning in myocardial 18F-FDG PET studies using both %ltered backprojection and iterative reconstruction. J Nucl 3. Weber WA. PET for response assessment in oncology: Med 2004;45:169-75. radiotherapy and chemotherapy. Br J Radiol 2005;78:42-9. 16. von Schulthess GK, Steinert HC, Hany TF. Integrated PET- 4. Weber WA. Chaperoning drug development with PET. CT: current applications and future directions. Radiology J Nucl Med 2006;47:735-7. 2006;238:405-22.

5. Cheson BD, P%stner B, Juweid ME, Gascoyne RD, Specht 17. Watson CC. New, faster, image-based scatter correction L, Horning SJ, et al. Revised response criteria for malignant for 3D PET. IEEE Trans Nucl Sci 2000;47:1587-94. lymphoma. J Clin Oncol 2007;25:579-86. 18. Anderson JM, Mair BA, Rao M, Wu CH. Weighted least- 6. Juweid ME, Stroobants S, Hoekstra OS, Mottaghy FM, squares reconstruction methods for positron emission Dietlein M, Guermazi A, et al. Use of positron emission to- tomography. IEEE Trans Med Imaging 1997;16:159-65. mography for response assessment of lymphoma: Consen- sus of the Imaging Subcommittee of International Harmo- 19. Browne J, de Pierro AB. A row-action alternative to the nization Project in lymphoma. J Clin Oncol 2007;25:571-8. EM algorithm for maximizing likelihood in emission tomography. IEEE Trans Med Imaging 1996;15:687-99. 7. Geus-Oei LF, van der Heijden HF, Corstens FH, Oyen WJ. Predictive and prognostic value of FDG-PET in nonsmall-cell 20. Boellaard R, van Lingen A, Lammertsma AA. Experimen- lung cancer: a systematic review. Cancer 2007;110:1654-64. tal and clinical evaluation of iterative reconstruction (OSEM) in dynamic PET: quantitative characteristics and e(ects on 8. Mawlawi O, Townsend DW. Multimodality imaging: an kinetic modeling. J Nucl Med 2001;42:808-17. update on PET-CT technology. Eur J Nucl Med Mol Imaging 2009;36 Suppl 1:S15-S29. 21. Westerterp M, Pruim J, Oyen W, Hoekstra O, Paans A, Visser E, et al. Quanti%cation of FDG PET studies using 9. Pichler BJ, Judenhofer MS, Wehrl HF. PET/MRI hybrid ima- standardised uptake values in multi-centre trials: e(ects of ging: devices and initial results. Eur Radiol 2008;18:1077-86. image reconstruction, resolution and ROI de%nition parameters. Eur J Nucl Med Mol Imaging 2007;34:392-404. 10. Beyer T, Antoch G, Muller S, Egelhof T, Freudenberg LS, Debatin J, et al. Acquisition protocol considerations for com- 22. Visvikis D, Cheze-LeRest C, Costa DC, Bomanji J, Gacino- bined PET-CT imaging. J Nucl Med 2004;45 Suppl 1:25S-35S. vic S, Ell PJ. In+uence of OSEM and segmented attenuation correction in the calculation of standardised uptake values 11. Beyer T, Antoch G, Bockisch A, Stattaus J. Optimized for [18F]FDG PET. Eur J Nucl Med 2001;28:1326-35. intravenous contrast administration for diagnostic whole- body 18F-FDG PET-CT. J Nucl Med 2005;46:429-35. 23. Defrise M, Kinahan PE, Townsend DW, Michel C, Si- bomana M, Newport DF. Exact and approximate rebin- 12. Townsend DW. Physical principles and technology of cli- ning algorithms for 3-D PET data. IEEE Trans Med Imaging nical PET imaging. Ann Acad Med Singapore 2004;33:133-45. 1997;16:145-58. 13. Lubberink M, Boellaard R, van der Weerdt AP, Visser FC, 24. Karp JS, Surti S, Daube-Witherspoon ME, Muehllehner Lammertsma AA. Quantitative comparison of analytic and G. Bene%t of time-of-+ight in PET: experimental and clinical iterative reconstruction methods in 2- and 3-dimensional results. J Nucl Med 2008;49:462-70. dynamic cardiac 18F-FDG PET. J Nucl Med 2004;45:2008-15.

52 Chapter 3: PET imaging instrumentation and principles of PET protocol optimisation

25. Brix G, Doll J, Bellemann ME, Trojan H, Haberkorn U, 34. Young H, Baum R, Cremerius U, Herholz K, Hoekstra O, Schmidlin P, et al. Use of scanner characteristics in iterative Lammertsma AA, et al. Measurement of clinical and subcli- image reconstruction for high-resolution positron emissi- nical tumour response using [18F]-+uorodeoxyglucose and on tomography studies of small animals. Eur J Nucl Med positron emission tomography: review and 1999 EORTC 1997;24:779-86. recommendations. European Organization for Research and Treatment of Cancer (EORTC) PET Study Group. Eur J 26. Mourik JE, Lubberink M, van Velden FH, Kloet RW, van Cancer 1999;35:1773-82. Berckel BN, Lammertsma AA, et al. In vivo validation of reconstruction-based resolution recovery for human brain 35. Boellaard R, Krak NC, Hoekstra OS, Lammertsma AA. studies. J Cereb Blood Flow Metab 2010;30:381-9. E(ects of noise, image resolution, and ROI de%nition on the accuracy of standard uptake values: a simulation study. J 27. Weber WA. Positron emission tomography as an ima- Nucl Med 2004;45:1519-27. ging biomarker. J Clin Oncol 2006;24:3282-92. 36. Krak NC, Boellaard R, Hoekstra OS, Twisk JW, Hoekstra CJ, 28. Avril N, Sassen S, Schmalfeldt B, Naehrig J, Rutke S, Lammertsma AA. E(ects of ROI de%nition and reconstruc- Weber WA, et al. Prediction of response to neoadjuvant tion method on quantitative outcome and applicability in chemotherapy by sequential F-18-+uorodeoxyglucose a response monitoring trial. Eur J Nucl Med Mol Imaging positron emission tomography in patients with advanced- 2005;32:294-301. EANM stage ovarian cancer. J Clin Oncol 2005;23:7445-53. 37. Stahl A, Ott K, Schwaiger M, Weber WA. Comparison 29. Borst GR, Belderbos JSA, Boellaard R, Comans EF, De of di(erent SUV-based methods for monitoring cytoto- Jaeger K, Lammertsma AA, et al. Standardised FDG uptake: a xic therapy with FDG PET. Eur J Nucl Med Mol Imaging prognostic factor for inoperable non-small cell lung cancer. 2004;31:1471-9. Eur J Cancer 2005;41:1533-41. 38. Thie JA. Understanding the standardized uptake va- 30. Boellaard R. Standards for PET image acquisition and lue, its methods, and implications for usage. J Nucl Med quantitative data analysis. J Nucl Med 2009;50 Suppl 1:11S- 2004;45:1431-4. 20S. 39. Weber WA. Use of PET for monitoring cancer therapy 31. Boellaard R, O’Doherty MJ, Weber WA, Mottaghy FM, and for predicting outcome. J Nucl Med 2005;46:983-95. Lonsdale MN, Stroobants SG, et al. FDG PET and PET-CT: EANM procedure guidelines for tumour PET imaging: ver- 40. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST sion 1.0. Eur J Nucl Med Mol Imaging 2010;37:181-200. to PERCIST: evolving considerations for PET response criteria in solid tumors. J Nucl Med 2009;50 Suppl 1:122S-150S. 32. Boellaard R, Oyen WJ, Hoekstra CJ, Hoekstra OS, Visser EP, Willemsen AT, et al. The Netherlands protocol for stan- 41. Jaskowiak CJ, Bianco JA, Perlman SB, Fine JP. In+uence of dardisation and quanti%cation of FDG whole body PET reconstruction iterations on F-18-FDG PET-CT standardized studies in multi-centre trials. Eur J Nucl Med Mol Imaging uptake values. J Nucl Med 2005;46(3):424-428. 2008;35:2320-33. 42. Visvikis D, Cheze-LeRest C, Costa DC, Bomanji J, Gacino- 33. Shankar LK, Ho(man JM, Bacharach S, Graham MM, vic S, Ell PJ. In+uence of OSEM and segmented attenuation Karp J, Lammertsma AA, et al. Consensus recommendations correction in the calculation of standardised uptake values for the use of F-18-FDG PET as an indicator of therapeutic for [18F]FDG PET. Eur J Nucl Med. 2001;28:1326-35. response in patients in national cancer institute trials. J Nucl Med 2006;47:1059-66. 43. Dai KS, Tai DY, Ho P, Chen CC, Peng WC, Chen ST, et al. Accuracy of the EasyTouch blood glucose self-monitoring system: a study of 516 cases. Clin Chim Acta. 2004;349:135-41.

53 Chapter 4: CT instrumentation and principles of CT protocol optimisation Ann Heathcote, Amy Wareing, Angela Meadows

Introduction To help understand the fundamental princi- This chapter commences with an overview ples of CT, knowledge of the basic CT imag- of computed tomography (CT) development ing system con%guration is required. Figure and an outline of the basic CT system con%gu- 1 identi%es the three main components: the ration. Image production is then discussed, CT scanner, the computers that control the focussing on image acquisition, reconstruc- scanner and the image display/image archive tion and post-processing. Thereafter, general aspects of the system. parameters and terminology for CT are high- lighted to support the remainder of the chap- Figure 2 demonstrates a typical PET-CT scan- ter, which addresses CT protocol optimisation ner. The two systems share the same housing, and attenuation correction in PET-CT. Key rel- with the CT scanner to the front and PET to evant pitfalls are considered which can lead to the rear. The central bore and surrounding degradation of the PET and CT image quality structures within the housing are referred to when best practice is not followed. It is recom- as the gantry. mended that if you have no prior knowledge of or background in CT, you should read the The gantry is a rotating framework that the recommended literature and references to patient moves through on the patient table gain greater insight into the subject. during data acquisition. It holds the x-ray tube, x-ray generator, slip rings, detectors, collima- The development of CT and basic tors and digital acquisition system (DAS) [1]. construction of a CT scanner The x-ray tube is responsible for the produc- CT scanning was invented by Geo(rey Houn- tion of x-ray photons. The %lters are respon- s%eld in the 1970s. In the beginning only one sible for removing low-energy x-ray photons, image was produced per rotation of the x-ray thereby reducing patient dose (to be detailed tube and image quality was consequently very later in this chapter). The collimators are used poor compared with the detail and resolution to de%ne the slice thickness and localise the achievable today. x-ray %eld to the area of interest. The detec- CT technology has developed signi%cantly tors capture the x-ray photons after they have over the last 20 years, with the advent of spiral passed through the patient and convert them CT in the 1990s and the subsequent introduc- ultimately into digital information via the DAS. tion %rst of dual-slice CT scanners and then of multi-slice scanners with the capability of generating 16, 64 and 128 slices per rotation.

54 Chapter 4: CT instrumentation and principles of CT protocol optimisation

Figure 1: A typical CT imaging system con%guration EANM

Figure 2: A typical PET-CT scanner Alliance Medical. © A. Heathcote

PET & CT scanner within shared housing

Gantry

Patient table © A. Meadows Alliance Medical.

55 During the examination, the patient lays on Acquisition the patient table. The table moves through the As the x-ray tube rotates around the patient, scanner gantry during the CT acquisition and the detectors measure the radiation transmit- subsequently moves further into position for ted through the patient from various loca- the PET scan to the rear of the housing. There tions (Fig. 3). The attenuation measurements is a weight limit for the table to ensure that it are calculated by the computer and stored as will move reliably during image acquisition. raw data %les (also called projections). Modern scanners collect projections from 360° and typ- The computer system that ‘controls’ the scan- ically measure 800-1500 projections per image. ner receives image data from the DAS and ap- plies a series of reconstruction algorithms to Attenuation is the reduction of the intensity of a produce a cross-sectional image. beam of radiation as it passes through an object – some photons are absorbed, others are scat- Image acquisition, reconstruction and tered. In CT, attenuation depends on the e(ective post-processing atomic density (atoms/vol), the atomic number The basic system con%guration has been out- of the absorber and the photon energy (Fig. 3). lined above; we shall now focus on image production, sub-categorised as image acquisition, The computer system receives the digital data reconstruction and post-processing. from the DAS and processes it to reconstruct the cross-sectional image. The computer system also enables general image techniques such as windowing, multi-planar reconstructions and 3D imaging [1]. Figure 3: The data acquisition process © A. Heathcote Alliance Medical. © A. Heathcote

56 Chapter 4: CT instrumentation and principles of CT protocol optimisation

Reconstruction The interpolation process is required since the The di(erential attenuation information is ac- calculation of images from a 360° spiral results in quired by the detectors and converted to a digi- image artefacts, as di(erent areas of the object are tal signal. The computer processes involved in measured at the start and end of the segment. image reconstruction include algorithms, convolution and interpolation. An algorithm is basically Convolution and back-projection a mathematical formula – in the CT scenario, The interpolated data then undergo the con- algorithms are used to reconstruct the image. volution-back-projection procedure to produce Convolution is a mathematical process applied the %nal CT image. In back-projection, each pro- to modify an image. Interpolation is speci%cally jection value that is produced is placed into the used in helical scanning and is a mathematical corresponding area in an imaging matrix. In process to calculate or estimate information simple back-projection, each projection value EANM based on known values of adjacent information. is added to all areas of the picture along the direction in which it has been acquired [2]. Interpolation Linear interpolation (Fig. 4) also known as The back-projection method on its own re- z-interpolation, allows the generation of a sults in an unsharp and inadequate image. consistent data set from the volume using an To remove the unsharpness, each projection arbitrary image position known as z . It en- is ‘convoluted’ prior to back-projection being R ables the linear interpolation of data measured done. It is important to note that convolution at a speci%ed angular position just before and and back-projection would warrant a chapter after a set table (slice) position (z ). in their own right, and we therefore recom- R mend further reading to study this concept in Figure 4: Diagram demonstrating interpolation more depth if further clari%cation is required.

The CT image CT is the measurement and ‘demonstration’ of the linear attenuation coe#cients µ(x,y) of the structures that the x-ray beam passes through during the examination. The value µ is converted into a CT value relative to the attenuation of water; this is to make the value more user friendly. The CT value/number is displayed in Houns%eld © A. Heathcote Alliance Medical. © A. Heathcote units (HU).

57 For a tissue (T) with an attenuation coe#cient Therefore, each individual tissue type, e.g. fat or µ , the CT value is de%ned as [2]: bone, is displayed with a di(erent HU/CT value. T Generally, the following CT values are accepted: CT value = (µ µ ) / µ × 1000 HU water = 0, air = -1000 and dense bone = +1000. T water water Figure 5 presents the Houns%eld scale and provides Figure 5: The Houns%eld scale further examplep values for various tissue densities. © A. Heathcote Alliance Medical. © A. Heathcote

Post-processing Post-processing is the term used to describe t 3D imaging – the generation of 3D images how the resultant CT image can be manipulated such as surface-shaded (SSD) reconstruc- by the operator. A wide variety of post-process- tions used in orthopaedic imaging and ing functions are available to the CT operator: maximum intensity projections (MIPs)

t Windowing – the manipulation of the contrast within the image to demonstrate certain structures or tissue types

t Multi-planar reconstructions (MPRs) – generation of 2D images in the sagittal/coronal/axial plane

58 Chapter 4: CT instrumentation and principles of CT protocol optimisation

Manipulation of the window width (WW) and Figure 6a,b: Images of liver metastases. Three window level (WL) is the most commonly used liver metastasis (1-3) can be seen using post-processing technique in CT [1]; WW is de- standard abdominal window settings of 400 %ned as the range of the CT numbers within WW and 40 WL (a) but a fourth metastasis (4, the image and WL is de%ned as the number at arrow) can be clearly identi%ed when using the centre/mid point of the range. For example, a narrow WW/WL setting of 250 WW and 40 when viewing the images, we select a WL setting WL (b) to represent the density of the tissue of interest, e.g. WL 40 = kidney. The WW determines the amount of contrast within the image set, e.g. WW = 400. Therefore, in this example the range 1 of levels demonstrated will be from -160 to +240. EANM It is important to note that the narrower the WW, the greater will be the contrast within the image. Figure 6 provides an example of how the choice 2 of settings can in+uence the imaging %ndings.

Upon completion of post-processing the CT 3 data, the resultant images will be archived. For

the purposes of PET, the data will have been Alliance Medical. © A. Heathcote used for attenuation correction of the PET data and subsequently the CT data will be presented as CT and fused PET-CT datasets to assist in anatomical localisation of pathologies (to be discussed later in the chapter). Ultimately, the image data can be archived onto either optical disc or CD or directly onto picture archiving communication systems (PACS).

4 © A. Heathcote Alliance Medical. © A. Heathcote

59 General parameters and terminology for CT CT protocol optimisation (dose reduction Below is a list of the common parameters and strategies) terminology that are used in CT. Awareness A number of methods and techniques are of these will be required for the remainder of available to minimise the radiation dose to a this chapter. patient undergoing a CT examination, including Auto mA, Smart mA, mA range limitation t Acquisition – process in which a single and optimum patient positioning: continuous set of scan data is acquired without a pause 1. Auto mA – mA modulation per rotation based on the last acquired scout and noise index se- t kV – penetrating power of x-ray lected by the user. Auto mA modulates the patient dose in the z-direction and ensures consis- t mA – tube current tent image quality independent of patient size. t mAs – mA × s The noise index value (NI) allows the user an absolute method to reproduce the same scan t s – time (seconds) at a later date. The NI represents the number of x-ray photons per rotation and can be user t Pitch – longitudinal distance (mm) that the de%ned; it is unique to the slice thickness and table moves during one rotation of the tube body part selected. Selection of the NI re+ects the level of noise acceptable to the radiologist t Pitch ratio (PR) – the pitch divided by the for a given CT examination. The CT scanner slice thickness, e.g. 1.5 or 1.5:1 will then automatically select, within a preset range, the tube current (mA) required to main- t Image index (II) – the distance (mm) be- tain the level of noise under the NI, taking into tween the centre of two consecutive slices account the patient’s attenuation. t Images per revolution – pitch divided by 2. Smart mA – mA modulation as the tube image index rotates around the patient. The mA table per rotation is based on the last acquired scout t Number of images – total number of im- and noise index selected. Smart mA can be ages reconstructed for one acquisition used in conjunction with Auto mA to further reduce patient dose. When using Smart mA, the beam is modulated four times to minimise the dose (Fig. 7).

60 Chapter 4: CT instrumentation and principles of CT protocol optimisation

Figure 7: Dose modulation when using Smart mA In CT, ‘bowtie’ %lters are used to reduce the surface dose by attenuating the unwanted x-rays at the edge of the beam (Fig. 8). Figure 8: The e(ect of the bowtie %lter © General Electric Company. © General Electric Company. All rights reserved.

3. mA range limitation – The mA exposure used can be ‘capped’ when a preset maximum level is reached. This has its limitations, however, in that EANM when the maximum mA is reached, the resultant image quality is adversely a(ected as the noise within the image is greater than that of the selected NI. As a result, consideration must be given to a compromise. For the purpose of © General Electric Company. All rights reserved. PET-CT only a low-dose image acquisition is It is important to ensure that the correct scan required as the intention is to acquire an image %eld of view (SFOV), i.e. bowtie %lter, is selected. more speci%cally for the purpose of attenuation For example, the head SFOV adds an extra algo- correction for the PET data (typically an mA of rithm to post-processing to reduce the beam- 80-110 is used as a capped value). As it is not hardening e(ect at the bone/brain interface. intended to produce images of ‘diagnostic’ CT This is a factor which should be considered value unless speci%cally requested, some image during imaging protocol management at the degradation is to be expected on the CT. initial protocol set-up for PET. Nonetheless, the technologist must understand how %lter selec- 4. Patient position – As PET-CT systems have tion and, more importantly, patient position the ability to vary the height of the table when within the gantry can impact on image quality the patient is within the gantry, it is essential and patient surface dose if the patient is not that the technologist knows the importance centered within the FOV. Figure 9 demonstrates of centralising the patient within the %eld of how patient dose can be adversely a(ected if view and understands the impact this can the patient is not correctly centered and how have on subsequent image quality and pa- images are degraded. Subsequently, the poor tient surface dose, particularly as %lters are quality CT data will in turn reduce the quality of used in CT in an e(ort to improve both factors. the attenuation-corrected PET data, for which

61 the CT scan is initially acquired. Consequently, If the patient is o( centre, with a %xed mA, the noise within the image is also increased by too little dose will reach the centre of the pa- mis-centering the patient (Fig. 10). tient as it will be %ltered out, and the body surface will receive excessive dose. More im- Figure 9: The e(ect of mis-centering on the portantly, when using Auto mA, if the patient patient dose is o( centre, the calculations for Auto mA will be incorrect. Calculations for mA based on attenuation are taken at the isocentre of the scanner. Therefore, good positioning and use of positioning lasers is critical to make best use of the %lters, thus improving image quality and reducing patient surface dose. © General Electric Company. © General Electric Company. All rights reserved.

62 Chapter 4: CT instrumentation and principles of CT protocol optimisation

Attenuation correction in PET-CT imaging e(ects are more severe for coincidence imag- Presuming that an adequate low-dose CT ing than for single-photon imaging (SPECT) acquisition for the purpose of PET has been [3]. Despite the photons being of a greater en- acquired, we now need to consider how these ergy than those used in SPECT, PET relies on data is used for the purpose of attenuation two photons simultaneously escaping across correction of the PET data. a greater mean photon path distance to be detected as a true event. The loss of true co- The process of correcting images for attenu- incidence event detection in PET is known to ation artefacts in nuclear medicine has been range from 50% upwards to as high as 90% in widely utilised for many years. Attenuation in large patients [4]. Loss of counts from attenu- nuclear medicine imaging is described as the ation correction (AC) increases image noise, loss of detection of true gamma ray events due image artefacts and image distortion (Fig. 11), EANM to either scattering out of the FOV of the detec- and these problems are magni%ed when imag- tor or absorption within the body. Attenuation ing larger patients (Fig. 12).

Figure 11: Comparison between non-AC and AC PET MIP images in a patient with an average BMI

Non-AC AC

Increased activity in areas that have low attenuation (E.g air %lled, decrease of absorption.)

Increased activity at body surface edge due to lack of attenuation at surface compared to less super%cial structures.

Areas of increased activity distort due to varying intensity of attenuation in di(erent directions (E.g. Bladder) (University of Virginia 2006). © A. Wareing Alliance Medical. Wareing © A.

63 Figure 12: Comparison between non-AC and AC PET coronal images of a patient with a high BMI

Non-AC AC

Pathology is Pathology is di#cult to easier to identify identify from from nearby nearby tissues. tissues. © A. Wareing Alliance Medical. Wareing © A.

Attenuation correction may or may not im- Almost all of the new sales for PET imaging are prove lesion detection, depending upon combined PET-CT scanners [6]. This is largely body habitus, distribution of FDG and so on. due to the practical and e(ective approach of Nonetheless, it will produce a more realistic acquiring co-registered anatomical and func- radioactivity distribution which is essential for tional images in a single scanning session. Fig- quanti%cation of tracer uptake; the values pro- ure 13 demonstrates PET with CT attenuation duced are referred to as standardised uptake correction. Much work has been conducted values (SUVs) [5]. indicating that fusion of PET and CT image data signi%cantly increases both sensitivity Methods of attenuation correction in PET and speci%city (see [7] for an early study). Historically, prior to the use of CT, a transmission scan was used to correct for attenuation of emission data in PET imaging. Early generation scanners utilised germanium-68/gadolinium-68 rotation rod sources to produce transmission images. Later generation scanners combined PET and CT in a single housing.

64 Chapter 4: CT instrumentation and principles of CT protocol optimisation

Figure 13: Example of PET with CT attenuation correction Top left: whole-body CT scout image; top middle: CT, axial slice showing lung pathology; top right: same axial slice, PET-only data. Below: fused axial slice demonstrating co-registration of functional and anatomical information

Image fusion of CT data and PET scan image. EANM © A. Wareing Alliance Medical. Alliance Medical. Wareing © A.

Attenuation correction and potential pitfalls photon energy (511 keV) [8]. The pixels are then At present, CT is the AC of choice. In addition segmented based on their HU and scaled ac- to the advantages of CT-AC discussed above, cordingly by various algorithms. Although most it is however important to recognise the po- applications using the very latest software can tential pitfalls of this method. Most oncology correct for discrepancies, there is a tendency patients referred for a PET examination have for AC algorithms to ‘over-correct’ occasionally. undergone a diagnostic CT examination with or without intravenous or ingested contrast Much literature has been produced discuss- solutions. Therefore, PET-CT examinations are ing CT-AC over-correction artefacts, including most commonly carried out using a low-dose IV contrast agents, oral contrast preparations, CT technique, which is often labelled ‘non-di- chemotherapy ports and dense/metallic ob- agnostic’. As a result, the potential for bias in CT- jects. A relatively simple and essential solu- AC can be high. This is because the measured tion for any interpreting physician is to assess CT Houns%eld unit (HU), related to the linear the non-AC PET images during the reporting attenuation seen by the x-ray beam, must be review; it is therefore essential that non-AC transformed to the corresponding higher PET data are always provided for interpretation [8].

65 Figure 15a,b: Patient arm position a(ecting AC Figure 14 provides two examples where image when using low-dose CT artefacts are apparent and streaking can be seen across the CT image, causing a potential for over-correction artefact. Clearly the pacemaker demonstrates a marked di(erence and is unavoidable. However, the metal button on a patient’s trousers could have been avoided with better patient preparation. Figure 14a,b: CT-AC and potential for over- correction. a A pacemaker causing streak artefact; b a metal button on a patient’s

trousers Alliance Medical. Wareing © A. 15a

14a © A. Wareing Alliance Medical. Wareing © A. 15b © A. Wareing Alliance Medical.Wareing © A. Alliance Medical. Wareing © A. 14b

66 Chapter 4: CT instrumentation and principles of CT protocol optimisation

Although artefacts from CT attenuation can oc- Conclusion cur due to poor positioning within the FOV (as In summary, an overview has been provided of previously discussed), equally scanning patients CT development, basic system con%guration with their arms by their side can result in beam- and the instrumentation and principles of CT. hardening/streak artefacts across the abdomen, More speci%cally, potential pitfalls have been where there is the greatest patient density (Fig. discussed, from patient position and prepara- 15). Therefore, good practice requires that pa- tion through to ‘over-correction’ complications. tients are scanned with their arms raised above It is essential that the technologist is aware of their head if at all possible. If the patient can raise the pitfalls to avoid prior to image acquisition, only one arm, then this is accepted as a compro- and that the image interpreter is also aware of mise. If the arms are required in the FOV as part the pitfalls and how they are presented when of a total body scan where the skin is the focus inevitable – particularly the importance of EANM of the study, they should be appropriately immo- making reference to non-AC image data. In bilised over the anterior aspect of the abdomen; concluding, we would once again emphasise this ensures they are within the FOV and thus that if you have little prior knowledge of or reduces the beam-hardening e(ect. background in CT, it would be advisable to read the suggested texts to gain a greater depth of knowledge on the subject.

67 References Chapter 4

References Suggested reading 1. Seeram E. Computed tomography – physical principles, Brink JA. PET/CT unplugged: the merging technologies of clinical applications and quality control. 3rd ed. St. Louis: PET and CT imaging. AJR Am J Roentgenol 2005;184:S135-7. Saunders; 2009. Costa DC, Visvikis D, Crosdale I, Pigden I, Townsend C, 2. Kalender WA. Computed tomography – fundamentals, Bomanji J, et al. Positron emission and computed X-ray system technology, image quality, applications. Erlangen: tomography: a coming together. Nucl Med Commun Publicis MCD Verlag; 2000. 2003;24:351-8.

3. Turkington TG. Attenuation correction in hybrid positron Kalender WA. Computed tomography – fundamentals, emission tomography. Semin Nucl Med 2000;30:2255-67. system technology, image quality, applications. Erlangen: Publicis MCD Verlag; 2000. 4. University of Virginia. Attenuation correction. Published by the Rector and Visitors of the University of Virginia. 2006. Kinahan P. CT-based attenuation correction for PET/CT scanners (ppt). Imaging Research Laboratory, Dept. of Radiology, http://www.med-ed.virginia.edu/courses/rad/PETCT/At- University of Washington. USA. 2005. tenuation.html http://depts.washington.edu/nucmed/IRL/ 5. Kamel E, Hany TF , Burger C , Treyer V , Lonn AH , von pims/2005_03_30/PETCTACv2r.pdf Schulthess GK , et al. CT vs 68Ge attenuation correction in a combined PET/CT system: evaluation of the e(ect of Seeram E. Computed tomography – physical principles, lowering the CT tube current. Eur J Nucl Med Mol Imaging clinical applications and quality control. 3rd ed. St. Louis: 2002;29:346-50. Saunders; 2009.

6. BIO-TECH Systems Inc. The market for PET radiophar- Wahl RL. Why nearly all PET of abdominal and pelvic cancers maceuticals and PET imaging. BIO-TECH Report # 300. Las will be performed as PET/CT. J Nucl Med 2004;45 Suppl Vegas: BIO-TECH Systems Inc; 2008. 1:82S-95S.

7. Beyer T, Townsend DW, Brun T, Kinahan PE, Charron M, Roddy R, et al. A combined PET/CT scanner for clinical oncology. J Nucl Med 2000;41:1369-79.

8. Mehta A, Mehta A, Laymon C, Blodgett CM. Calci%ed lymph nodes causing clinically relevant attenuation correction artifacts on PET/CT imaging. J Radiol Case Reports 2010;4:31-7.

68 control (QC) procedures. A successful QA pro- quality appropriate includes which place in be to programme (QA) assurance quality a forrequirement legal a is there PET-CT using As with all medical imaging equipment, when Introduction Peter Julyan for PET-CT control assuranceandquality Chapter 5:Quality tient preparation, may be just as important important as just be may preparation, tient pa- as such article, this of scope the beyond explicitly aspects some that noted be to is It mode andpredominantly inoncology. rodeoxyglucose (FDG), usually in whole-body tracer of glucose metabolism, +uorine-18 +uo- trapped the imaging involve investigations most addition, In scanner. CT x-ray an with scanner PET a combines equipment current all virtually that recognised is it so, doing In equipment, concentrating on the PET aspects. strumentation QA and QC with current PET-CT infor requirements the outlines article This imaging.ture ofPET of CT, and secondly the more quantitative na- with two additional factors: %rstly the inclusion for any nuclear medicine imaging equipment, as same the are principles general the tems, sys- PET-CT Forpromptly. faults identifying system uptime and improve image quality by resultsCarefulincreaseof should monitoring been performed and their results, monitored. have tests appropriate that demonstrates it a necessary part of any QA programme in that malfunctioning equipment. Record keeping is with patients scanning of likelihood the ing gramme will reduce image artefacts by reduc- 69 terise the equipment (as may be suitable for suitable be may (as equipment the terise of measureperformance data to fully charac- togenerally producenot is exhaustivean set here [3]. goal standard The 2-2007 NU NEMA current the follow should which measures, the system and conduct a set of performance up set correctly to installation at done work the with commences programme QA Any precision (typically with a long, low count-rate high to measured be then will response of tings. As with gamma cameras, the uniformity set- optimal from drift for potential is there and necessary is calibration of level other an- manufacturers, major the from scanners capable time-of-+ight of emergence recent the Withrequired. then is calibration timing a events coincident For elements. detector individual identify correctly to used then are an appropriate energy window. These signals within photons 511-keV for signals propriate togivephotomultipliertubes the of ap- gain bration. Fundamentally, calibration will set the cali- regular and visits service periodic clude in- will programme service subsequent The a baselinefor follow-up measurements. parameters from the tender process and act as performance the con%rm to but equipment) particular that on publication benchmark a QA programme obtained elsewhere [e.g. 1,2]. be can aspects these on information gation; mation obtained investi-through the PET-CT infor-diagnostic the of quality overall the for

EANM acquisition) and is used in data reconstruc- rect half-life) or a pre-corrected +at line. Either tion – this is often termed the normalisation way, such a test enables the user to under- for PET systems. stand the data from the scanner. For a gated acquisition (whether cardiac or respiratory) it A vital part of the calibration of a PET system is may sound redundant to perform a phantom the absolute calibration such that results may scan of a static source, but this will check that be expressed quantitatively in terms of kBq/ equal sensitivity is given to each bin. ml. For this purpose, a known source generally of 18 F is accurately measured in a dose calibra- More extensive testing may be appropriate tor. (For historical reasons this may be referred annually, when a full assessment of the CT to as the well counter calibration.) This will be performance should be undertaken and the discussed in more detail later. user may wish to repeat some of the initial NEMA performance measures, perhaps most The frequency of repeat calibration should be usefully the image quality phantom. guided by the manufacturer’s recommendations and will vary from system to system. PET daily QC The basis for PET scanner QC is the daily blank Regardless of the mechanism of calibration, sinogram formed by irradiating the detectors validation of the scanner must re+ect the approximately uniformly with 511-keV pho- intended range of studies to be performed. tons from a long-lived positron source, typi- Thus, if whole-body acquisitions are to be per- cally a rotating 68 Ge rod source within the gan- formed, whole-body acquisitions of 18 F or 68 Ge try or a 68 Ge cylinder placed within the %eld phantom(s) must be carried out, checking for of view. This is analogous to a uniform 57 Co example that suitable overlap between bed +ood as would be used for gamma camera positions has been set. Similarly, if dynamic QC. As there is not a one-to-one relationship acquisitions are to be performed, a representa- between the sinograms and the detector ele- tive dynamic validation should be carried out. ments, the data are often re-sorted into fan- Such validation can be done, for example, with sums, as though the whole detector ring had a decaying source of 18 F with dynamic framing been opened up and laid +at. This is illustrated typical of the intended frame durations; this in Fig. 1, where two sinograms are shown at enables one to check that, for example, the the top with the fansum for the entire detector scatter correction is robust down to the short- assembly at the bottom. Individual detector est intended frame duration. Analysis of such elements identi%ed as yellow and green points acquisition will give the user either a decaying correspond to lines in the sinograms. curve (which should be %tted to give the cor-

70 Chapter 5: Quality assurance and quality control for PET-CT EANM Courtesy of The Christie NHS Foundation Trust ChristieThe NHS Foundation Courtesy of Figure 1: Daily QC sinogram and fansum images (resulting from irradiation with a 68 Ge cylinder in the centre of the %eld of view on a Siemens system)

In general the manufacturer’s software will ing very few counts and required attention. automatically compare the current readings Exactly when errors seen in daily QC measure- with a reference set – typically acquired at the ments give rise to signi%cant problems is not last service visit – but the user should be aware easy to de%ne, which is why users may prefer of how these are generated. In the example to study reconstructed images rather than the in Fig. 2, one of the detector blocks was giv - somewhat less direct sinogram or fansum.

71 Daily QC Good

Daily QC – Block Error Courtesy of The Christie NHS Foundation Trust ChristieThe NHS Foundation Courtesy of

Figure 2: Daily QC fansum images for various detector parameters for good performance (above) and with a block error (below) (resulting from irradiation with an internal 68 Ge rod on a GE system)

When a central 68 Ge cylinder is used, this Quantitative PET data set may be employed to reconstruct There is the additional requirement in PET an image which may be analysed to give that the images produced should invariably an image set of the appropriate activity be fully quantitative, with uptake in each voxel level and uniformity in line with recent being expressed in terms of kBq/ml. On this measurements. Indeed, such a daily check basis, the standardised uptake value, SUV, is of quantitative accuracy can be very useful calculated by normalising for the administered and such measurement (with either 68 Ge or activity and patient weight (or a variant there- 18 F) may be required for clinical trials where of). This provides more re%ned diagnostic in- quanti%cation of the data is important. formation but also requires additional quality

72 Chapter 5: Quality assurance and quality control for PET-CT

considerations. While the following may seem here the periodic change between daylight obvious, careful adherence is vital and sites saving times needs to be carefully observed. have been found wanting in this respect [4, 5]. Correction for the residual activity post in- Ultimately the calibration of the scanner is jection must be applied, but this is compli- tied (via the scanner calibration process) to cated by the rapid decay of 18 F, necessitating the radioactive dose calibrator upon which thoughtful application of decay corrections. injections are measured. There is quite rightly the requirement that this calibrator must be For repeat imaging in the same patient, as accurate and traceable to national primary many factors as possible must be kept con- standards so that patient injected activities stant. On each occasion it must be ensured and therefore e(ective doses are as intended. that the patient is in the fasting state by check- EANM Nevertheless, in terms of calculating SUVs, a ing blood glucose levels. Consistency in the dose calibrator error would lead to a wrong timing of the emission measurement is also scanner calibration and these errors would important, as even at between 60 and 90 min cancel each other out, leading to correct SUVs. after injection there can still be appreciable It is important, though, to realise that the scan- changes in the SUV. ner is directly tied to a speci%c dose calibrator and there may be departments where small The %nal and crucial part in the generation of errors between calibrators (acceptable within results is the analysis software. It should also 5% of the national standard, say) could intro- be recognised that for the calculation of SUVs duce errors approaching 10% (with the cali- this software should be checked to give cor- brator o( in di(erent directions). rect results, using it in exactly the same way as is intended to be done for patients. Just as important in the calculation of SUVs is the patient weight, and patient weighing CT QA/QC scales must accordingly also be of suitable While the CT component of PET-CT is often certi%ed quality. used as fairly modest quality CT for the purpose of attenuation correction and anatomical The %nal element in the calculation of SUVs localisation of PET abnormalities, the require- lies in the relative timing of the injection ments for QA/QC are the same. It is neces- measurement and scanning. Nowadays the sary to have a QA programme in place that ready availability of radio-controlled clocks at recognises the need to produce appropriate a(ordable prices o(ers a practical means of quality CT images at the intended patient ef- establishing the correct time although even fective dose [6].

73 An annual check by experts in diagnostic The inclusion of a full PET-CT acquisition in radiology equipment testing will assess the the daily or weekly QC has the advantage of general radiation safety, CTDI (computed to- testing the full patient acquisition including mography dose index), slice width, absolute the database, table movement and alignment CT number, noise and resolution. of the PET and CT data sets.

A useful and simple daily or weekly test of For clinical trials, there may be a requirement the CT should be undertaken to ensure that to send o( data for pooled analysis across it gives the correct value (in Houns%eld Units, many sites. This should always be done with HU) for water, i.e. ~0 HU, with consistent noise phantom data prior to imaging the %rst pa- levels. tient, and all the steps must be followed that would be performed on patient data. For ex- Combined PET-CT QA/QC ample, if it is necessary to use a speci%c meth- The fundamental requirement of the com- od to anonymise patient images, this method bined PET-CT data sets is that they are cor- should be applied to the phantom data, too. rectly spatially aligned. The manufacturer will have procedures to set the physical alignment There are additional requirements if the PET- of the separate gantries to a certain tolerance CT is to be used for radiotherapy planning. together with acquisition of a speci%c test ob- Firstly, patient positioning must aim to be as ject to measure and correct for any residual close as possible to that which will be used misalignment. It is recommended that as well during the radiotherapy, with appropriate pa- as having an awareness of these operations, tient supports and an appropriate room laser users should consider independent tests. A alignment system. The integrity of the PET-CT combined PET-CT acquisition of an 18 F or 68 Ge in its transfer to the radiotherapy planning sys- cylinder is a simple form of such a test. More tem is also vital. If the CT portion is to be used sophisticated tests using small points %lled in the generation of radiotherapy plans, there with a mixture of radioactivity and contrast will be additional and possibly more stringent medium may also be undertaken and may requirements in terms of bed de+ections and include more extensive movement of a fully absolute calibration of the CT numbers. Most loaded bed. important is the close cooperation among various sta( groups in nuclear medicine and radiotherapy departments.

74 measurements ofpositron emissiontomographs. Performance 2-2007. NU Publication Standards NEMA 3. 2010;37:181–200. Imaging Mol sion 1.0.Eur JNuclMed ver-imaging: PET procedurefortumour EANM guidelines PET/CT: and PET FDG al. et SG, Stroobants MN, Lonsdale FM, Mottaghy WA,WeberMJ, O’Doherty R, Boellaard 2. with HD, Siegel BA, et al. Procedure guideline for tumor imaging Delbeke D, Coleman RE, Guiberteau MJ, Brown ML, 1. Royal References performance testing ofdiagnostic x-rayperformance systems. 2005 routine for standards Recommended 91. Report IPEM 6. imaging network experience. J Nucl Med 2009;50:1187-93. Radiology of College American the trials: clinical cancer gel BA. Quali%cation of PET scanners for use in multicenter Sie- AM, Levering JS, Karp JR, Sa(er JS, Scheuermann 5. 2007;34:392–404.meters. Imaging Mol Eur JNuclMed para- de%nition ROI and resolution reconstruction, image standardised uptake values in multi-centre trials: e(ects of using studies PET FDG of Quanti%cation al. et E, Visser A, Paans O, Hoekstra W, Oyen J, Pruim M, Westerterp 4. References Chapter 5 18 F-FDG PET/CT 1.0. J Nucl Med 2006;47:885-95. 1.0.JNuclMed F-FDG PET/CT 75 regulations. Heidelberg New Berlin York: 2005. Springer; and chemistry, physics, imaging: PET of Basics GB. Saha berg New York: 2005. Springer; sitron emission tomography: basic sciences. Berlin Heidel- Po- editors. MN, Maisey TownsendDL, DW,PE, Bailey Valk Suggested reading

EANM Chapter 6: PET isotope production Katy Szczepura

Introduction not its speed. This means that when a charged Radioisotope production for PET is generally particle enters a magnetic %eld, it will start performed by means of a cyclotron that is to travel in a circle. The faster the particle is used to accelerate charged particles. These travelling, the bigger the circle it will travel accelerated particles then go on to interact in. A cyclotron takes advantage of these two with a target to produce radioisotopes suit- phenomena and utilises them to accelerate able for use in PET imaging. positively charged particles.

This chapter will introduce the general prin- A cyclotron consists of two semi-circular con- ciples of operation of a cyclotron, explain how ducting structures known as dees, with an in- the particles interact in the target and discuss sulating gap between them. These dees are the production of some of the isotopes com- placed between two magnets with opposite monly used in clinical PET interactions. It will poles facing each other, so there is a magnetic also look at the general safety considerations %eld travelling from top to bottom. As the involved in the use of cyclotrons. charged particles enter the magnetic %eld, they will travel in a circular motion around the dees. The cyclotron A cyclotron is a type of particle accelerator that Once the charged particles are travelling in accelerates charged particles, such as protons a circular motion, there needs to be a way of and deuterons, to high energies. Before dis- accelerating them. An electric %eld is placed cussing how the cyclotron works, it is impor- between the surfaces of the two dees, such tant to understand how these particles behave that when the charged ion exits one dee, it will in the presence of electric and magnetic %elds. be repelled by the oppositely charged surface of that dee and attracted to the surface of the When a charged particle is in the presence second dee. This causes the particle to accel- of an electric %eld, it will feel a force that will erate and gain energy. As the particle is now accelerate it in the direction of the %eld. If this travelling at a faster speed, it will move in a acceleration is in the direction that the particle larger circle within the second dee. is already travelling in, then it will cause the particle to gain energy. When the particle reaches the surface of the second dee, it needs to be accelerated again, When a charged particle is in the presence and so the surface of the second dee needs to of a magnetic %eld, it feels a force that is per- become oppositely charged while the other pendicular to its direction of motion. This force surface needs to become charged to attract the will make the particle change its direction, but particle towards it, creating further accelera-

76 Chapter 6: PET isotope production

tion. This means that the direction of the elec- t This causes the particle to accelerate. tric %eld needs to change just as the particle emerges between the dees. This is achieved by t The particle moves in a circle when in a dee applying a high-frequency alternating voltage (the circle is larger due to higher speed) across the dee electrodes. The dees themselves are isolated, and so the particles are not a(ect- This process is repeated until the particle has ed by the electric %eld once they are inside. accelerated su#ciently, and so is travelling in a large enough circle, to be released from So, a summary of this process is: the dees. As this particle is now accelerated, it has gained energy. This energy is then used t The particle moves in a circle when in a dee. to interact within a target to create radioisotopes (Fig. 1). EANM t When it reaches the surface of the dee, it is attracted into the opposite dee due to the electric %eld.

Figure 1: Diagram illustrating the operation of a cyclotron (magnetic %eld oriented perpendicular to the dees, not shown) © K. Szczepura.

77 Ion source While many di(erent radioactive isotopes can For these events to occur there has to be a be produced in the cyclotron, in order to be source of ions, and the ion source is depen- suitable for PET imaging they must have the dent on the isotope that is being produced, as following properties: di(erent isotopes need di(erent interactions between target and particle. t Emit positrons when they decay

The ion source is a small chamber in the centre t Have an appropriate half-life of the cyclotron that produces either negative or positive ions, depending on the con%gura- t Be capable of being synthesised into a tion. These particles are attracted into the dees pharmaceutical to produce a useful tracer by electrostatic attraction. for studies in humans

Negative hydrogen ions (H -) are produced by When targets of stable elements are irradiated using a tungsten %lament to ionise hydrogen by placing them in the beam of accelerated gas. These particles are attracted into the dees particles, the particles interact with the nuclei by electrostatic attraction. The electrons are within the target and nuclear reactions take stripped o( the H - particle using a carbon place. foil, leaving an accelerated proton to interact within the target. A simple nuclear reaction induced by a proton A p, on a target Z X can be given by: The exiting charged particles are directed to the required target using a deviating electro- magnet. This means that di(erent isotopes can be produced using the same cyclotron A reaction induced by a deuteron d (a proton A depending on the target used. and a neutron), on a target Z X can be given by: Positron emitter production Once charged particles exit the cyclotron, they can go on to produce positron emitters by interacting with a target. The isotope that is produced depends on the typed of charged particle that has been aceelerated and the material from which the target is made.

78 Chapter 6: PET isotope production

Fluorine-18 18 F is produced by proton bombardment of oxygen-18-enriched water. The proton interacts with the 18 O and produces a neutron and 18 F.

Oxygen-15 15 O is produced by deuteron bombardment of natural nitrogen. The deuteron interacts with the 14 N and produces 15 O. EANM

Carbon-11 11 C is produced by proton bombardment of natural nitrogen. The proton interacts with the 14 N and produces a neutron and 11 C.

Nitrogen-13 13 N is produced by proton bombardment of distilled water. The proton interacts with the 16 O and produces an alpha particle and 13 N. © K. Szczepura.

79 Radiation protection issues when using a There are two ways of providing this shield- cyclotron ing. One is vault shielding, where the cyclotron Due to the radiation protection issues involved is housed in a protected room with concrete in the production and disposal of radioactive walls. The second is incorporation of the shield- materials, radiotracer production and cyclo- ing in the cyclotron, which is referred to as “self- tron operation are mostly automated. This is shielded”. In this option the steel frame of the normally achieved via a computer-controlled cyclotron provides the primary shielding, with menu that the operator employs to select the concrete blocks that are hydraulically driven isotope for use. providing complete radiation protection. The advantages of self-shielded cyclotrons are (a) Cyclotrons also need a lot of internal shielding they have a smaller footprint and so require less to protect sta( from high radiation doses, not space and (b) there are fewer decommissioning only from the positron emitters but also from implications. Currently, however, the cost of a the by-products that are produced, such as self-shielded system is almost the same as the the alpha particle in 11 C production. cost of building the concrete vault.

80 as culture, religion, age and pathology. Whilst dividual variations in+uenced by factors such in- re+ecting types patient of range broad a on the other they must care appropriately for arethat sponsibilities technical naturein and re- have they hand one the roles.On their to foci di(erent quite two having of virtue by position unique a in themselves %nd raphers radiog- and technologists medicine Nuclear Introduction Simona Cola andPeter Hogg Chapter 7:Patient care inPET-CT the nurse and others in looking after patients. dimension to patient care and complements or nuclear medicine physician, adds a di(erent The medical practitioner, such as a radiologist management. and care in ability heightened their of because bene%ts signi%cant genders nurse, but the inclusion of this professional en- management skills. Not all centres will have a and care patient of range a have to pected ex- be will sta( clinical the of most invariably but cyclotrons) operate that teams technical not be required to have patient care skills (e.g. will sta( non-clinical the speaking, Generally include both clinical and groups.non-clinical quitelargeand be centres can some teamin PET-CT. healthcare Here the multidisciplinary and medicine nuclear within true ticularly par- is this and alone operate not do raphers radiog- and technologists medicine Nuclear agement competencies. requirement to include patient care and man- the is them among exists that thread clear a countries in terms of emphasis and outcome, educational curricula vary between and within 81 regarding where the speci%c responsibilities speci%c the where regarding countries between exist variations but vice ser- radionuclide clinical the for responsible a suitably quali%ed medical practitioner will be Regulatory requirements normally dictate that part. their play should professionals clinical all that approachand team a entail agement man- and care patient that is stage this at about ourselves remind to point important an Perhaps management. and care patient bring their unique set speci%c toskill bear on they and technologist, medicine nuclear the In this whole context sit the radiographer and with that in mind we list three nursing texts texts nursing three list we mind in that with and evidence of body that reviewto chapter context of health. is It not the purpose of this the in people of care the with itself cerns con- that literature of range broad a is There PET-CT ment ofpatients. manage- and careappropriate the ambition: same the have both favoured, is alternatives these of whichever Nonetheless, standard. correct the to is conduct professional sures as- that body legal recognised nationally a to accountable are they Instead practitioner. medical that to accountable not and actions their for responsible legally and personally are professionals healthcare all component, unit, whilst in other countries, beyond the PET PET-CT a of charge ultimate in is practitioner some countries make it clear that the medical instance, For se. per experience PET-CT the lie for patient care and management outwith

EANM under suggested reading that may be of value the explanation will be tailored to his or her in this respect. Instead of giving a broad back- requirements. It is important to ask the patient ground to patient care and management, we whether they have understood the explana- shall focus purely on some matters that apply tion and whether they are happy to proceed. speci%cally to PET-CT. Obtaining informed consent from the patient, however, may be a matter for the medical Often, patients who attend for PET-CT exami- practitioner or another healthcare worker. nations are worried about their health and Whichever is the case, national legislation and this can manifest itself in anxiety that may be guidelines must be followed where they exist. evident on their arrival at the PET-CT centre. Various novel and common strategies [1, 2] to Figure 1: Patient information for PET-CT minimise patient anxiety have been described examinations: S.Maria Nuova Hospital’s in the literature, and emphasis is placed on booklet, Reggio Emilia, Italy the provision of information prior to, during and after clinical procedures [3]. Prior to the examination, letters, websites and information lea+ets have proved helpful in explaining what the PET-CT procedure involves. Similarly, if post-procedure information is required then these written forms of information are helpful. When constructing such information sources, it is essential that the information is conveyed in a fashion that the majority of the population will understand, and for certain languages readability checks can be used to assess this. Obviously, all patients are individuals, but certain patient categories may require special materials just for them (e.g. children). Often each PET-CT centre has its own information booklet or set of lea+ets; an example is shown in Fig. 1. On arrival at the PET-CT centre, the patient will be given a verbal explanation of the procedure; this will use vernacular that Courtesy of Department Emilia Reggio S. Maria of Nuclear Medicine, Hospital, Nuovo the patient will understand. Consequently each patient will be treated individually and

82 Chapter 7: Patient care in PET-CT

The areas in which the radiographer or nuclear Patient acceptance medicine technologist will have involvement Patient identity must be checked in accor- with patient care and management are as dance with any local policy that is in place. follows: The patient is normally asked for three forms of identi%cation, for example, name, date of t reception birth and address. The request form must also be checked to ensure that information on it t involvement of patients and/or other rel- conforms with that given by the patient. This evant people: patient empowerment procedure is necessary to minimise the risk of identi%cation error and misadministration of t communication and information the radioactive substances and CT exposure. When questioning the patient it is important EANM t comfort prior to, during and after the pro- to preserve their privacy and security. During cedure this aspect of the patient experience it is quite appropriate to give some preliminary informa- t safety tion about the PET-CT scan, including how long it will take and general details about the t privacy actual procedure.

Also they will have involvement with the fol- Patient preparation lowing %ve phases: If there are no contraindications to the PET-CT procedure then the patient can be prepared. A t patient acceptance simple but detailed explanation of the whole procedure should be given, making sure that t patient preparation the patient has understood what is required from them and what the procedure entails. t explanation of the procedure to patients Any relevant risks should be articulated, in ac- and relevant persons cordance with local policy and to minimise clinical negligence claims. For the PET-CT t patient comfort, safety and privacy examination the patient must stay in a quiet relaxing waiting room before and after the t monitoring of patient’s status during the injection. In this rest room there should be PET-CT examination comfortable waiting conditions with a suitable ambient temperature (Fig. 2). In certain instances, relaxing music might be played. The

83 correct amount of PET radiopharmaceutical and patient education can increase patient should be prepared, in line with any national motivation to comply; such upfront informa- or international guidelines that are being fol- tion can improve the patient experience and lowed. Similarly, it should be administered in also improve the diagnostic quality of the scan line with national or international guidelines. (e.g. they may move less because they know Finally, the administration should be docu- what to expect). The nature of any interaction mented appropriately. Further information with the patient will depend on the patient’s about the patient and radiation risk can be requirements; determinants for these require- found in Chap. 7 on radiation protection. ments may include patient baseline knowledge and understanding and the quantity and Figure 2: Hot waiting area where the patient type of information that needs to be imparted relaxes post injection to them. The latter is an interesting point because it is well known that not all patients wish to have a detailed explanation; when patients indicate that they want only basic information then that request should be granted to them, thereby protecting their human rights. As noted earlier, any explanation should use terminology which is consistent with the patient’s intellectual and subject-speci%c ability, and the use of technical terms may not always be appropriate. Courtesy of Department of Nuclear Medicine, Courtesy of Department of Nuclear Medicine, Emilia Reggio S. Maria Hospital, Nuovo

Explanation of the procedure to patients Patient comfort, safety and privacy and relevant persons Communication in this context may be de- Comfort %ned as the transfer of information from the During the PET-CT examination it is impor- healthcare worker to the patient and vice tant to use immobilization devices to avoid versa with a view to changing understand- patient movement but it is also necessary to ing and perception in the recipient. Radiog- use devices to improve patient comfort. De- raphers and nuclear medicine technologists vices that can be used to aid immobilisation should have well-developed communication and comfort include arm rests, knee rests and abilities and these should be used e(ectively a warm blanket. Figure 3 gives an indication to alleviate patient anxiety whilst maximising of how patients’ arms and legs can be made patient compliance. E(ective communication more comfortable.

84 Chapter 7: Patient care in PET-CT

Figure 3a,b: Arm and leg placement to ensure comfort during a PET-CT examination Courtesy of Department of Nuclear Medicine, Courtesy of Department of Nuclear Medicine, Emilia Reggio S. Maria Hospital, Nuovo

a b EANM Safety Privacy At all stages during the PET-CT procedure, the Privacy of personal information is governed radiographer or nuclear medicine technologist by national law and therein security of patient must take responsibility for ensuring that the data must be maintained. Many hospitals have patient’s physical well-being is optimal. Amongst speci%c data protection policies and these other things this involves adhering to medicine should be followed to the letter. Some hos- management policies in the event of the pa- pitals have a named individual who can be tient requiring drugs and gasses (e.g. oxygen). approached by sta( for advice and informa- Particular attention should be given to moving tion about local data protection policies and and handling of patients, and again it is impor- the law generally. In some countries, infringe- tant to adhere to local policies. Compliance with ment of the local data protection policy (and such policies heightens patient safety and also therefore the law) may be deemed both a civil that of the healthcare team (e.g. by minimising and a criminal o(ence and for the latter a jail the chance of a back injury). Patients should be sentence may be imposed. Aside from the observed at all times during the scan, whether legalities, the patient should be a(orded an through lead glass or video camera. Patients at a appropriate level of privacy, which is particu- high risk of injuring themselves, perhaps through larly important when they need to undress frailty, should be monitored closely prior to and and during communication of information of after the scan. Risk assessment procedures a personal and intimate nature. should have been conducted and be up-to- date, and policies arising from these assessments should be implemented in routine practice.

85 Monitoring the patient’s clinical status Summary during the PET-CT examination Patient care is a critical aspect of the radiog- Appropriate patient care will involve recog- rapher’s and nuclear medicine technologist’s nising and then responding appropriately to role. Patient care and management has been emergency situations. Various levels of emer- extensively studied and is well reported in the gency exist, from quite simple (e.g. faint) to se- nursing literature and you are encouraged to vere (e.g. heart failure). For PET-CT the use of x- access that material. Care and management of ray contrast media does give rise to reactions [4] the patient is a team approach and understand- and it is essential that the radiographer and the ing the role of other healthcare professionals nuclear medicine technologist have a thorough in that team is important. Radiographers and understanding of contraindications and drug nuclear medicine technologists have particular incompatibilities prior to their administration care and management responsibilities within and also of reactions post administration. The their role and they should discharge them in a radiographer and nuclear medicine technolo- competent and professional manner. gist must be adequately trained to recognise and deal with a broad range of emergency Simona Cola would like to thank all the Nuclear situations, and their competence to practice Medicine team of S. Maria Nuova Hospital Reggio should be updated in line with local policy. Emilia for their help towards this chapter. At the very least the training should involve a range of basic skills and also the ability to know when and how to call for help.

86 randomized Oncol2003;91:213-7. trial. Gynecol a colposcopy: undergoing patients for anxiety reduce to Chan Y,P,Lee music of Ng H, T,use WongL. Ngan The 2. care situation.JPers Psychol Soc 1983;44:1284-6. interpersonal impacts, and adjustment to a stressful health Auerbach S, Martelli M, Mercuri L. Anxiety, information, 1. References References Chapter 7 and review of the literature. Eur J Radiol Extra 2007;61:129-33. after non-ionic X-ray contrast medium injection: case report Böhm I, Schild H. Immediate and non-immediate reaction 4. Spine 2007;74:353-7. Bone Joint procedures, +uoroscopy-guided undergoing about interventional rheumatology: A study in 119 patients information of impact and Quality A. Saraux A, Roudaut V, Devauchelle S, Jousse-Joulin I, Samjee X, Guennoc 3. 87 2003;31:222-9. medicine technologist (2003 revision). J Nucl Med Technol nuclear the for guidelines responsibility and Performance speci#cPET-CT 8th ed. Philadelphia: Lippincott, Williams & 2008.Wilkins; methods. and research:principles Nursing C. PolitD, Beck York: 2008. Hill; McGraw ciples and evidence for practice. 2nd ed. Maidenhead New Payne S, Seymore J, Ingleton C. Palliative care nursing: prin- 2nd ed. New Basingstoke York: Palgrave 2006. MacMillan; Parahoo K. Nursing research: principles, process and issues. General patient care Suggested reading ments_physis_techs.pdf downloads/documents/06_04_-_pet-ct_training_require- [06/04] 2006 September 26 Approved Board Advisory PET-CT UK 1.1 Version technologists/radiographers. clinical for and scientist clinical for competencies scanning PET-CT http://www.bnms.org.uk/~bnms/images/stories/

EANM Chapter 8: Radiographer and technologist competencies – education and training in PET-CT Peter Hogg and Angela Meadows

Introduction that training has occurred – this might be post- This chapter commences with consideration graduate or post-basic. One thing is certain of where radiographer and nuclear medicine – one size will not %t all, principally because technologist PET-CT training and education of di(erences in the context of each country. might occur; it then progresses to the detail of which subjects might be learnt and the Formative professional training for radiographers competencies that should be obtained. Em- and technologists varies considerably between phasis is placed upon %rst post competence countries. For instance, some countries o(er in PET-CT, and due regard will be paid to the 2-year hospital-based certi%cates, while others requirements for practising to a level %t for have 3- or 4-year university-based bachelor de- purpose. To assist us in bringing together this grees and at least one o(ers a Masters of Science chapter we have drawn upon national guide- route. The decision to include PET-CT compe- lines [1, 2] produced within the United States tencies within formative professional education and also within the United Kingdom. We rec- should be well thought through, and this would ommend both of these documents to you. likely be re+ected in whether the %rst post of the At the end of the chapter we have indicated professional would have a high probability of in- some suggested reading; these texts focus on volving routine working within a PET-CT centre. the important areas of competence, accredita- If this is not the case (i.e. if, on quali%cation, pro- tion of prior learning and curriculum develop- fessionals are likely not to work within PET-CT) ment. If you are not familiar with educational then the inclusion of PET-CT within the forma- processes then we strongly recommend that tive professional curriculum may be of general you consider reading a range of similar edu- interest but the required several weeks of clinical cational texts prior to engaging in the design competence-based training might represent a of a PET-CT curriculum. poor investment. Of course, for radiographers this same argument could equally be applied When should PET-CT training occur? to ultrasound, computed tomography, magnetic Di(erent countries have di(erent models for resonance imaging, gamma camera nuclear training their radiographers and nuclear medi- medicine, interventional procedures and so on. cine technologists and even within the same For the purpose of ‘general interest’, should there country di(erent models can exist between be a desire to include background information these professional groups. Whichever group is on PET-CT within formative professional educa- considered, it is important to have a rationale tion then this might be better done within the for when PET-CT training should occur. Two classroom with a small amount of time spent options presently exist: the %rst is within forma- observing PET-CT in the clinical environment, tive professional training; the second is after so as to reinforce the theory. This latter option

88 Chapter 8: Radiographer and technologist competencies – education and training in PET-CT

would not be intended to imply competence to formative professional education then two quite practice PET-CT. Should the requirement for PET- di(erent educational models exist for students CT competence to practice be opted for within to negotiate (Fig. 1) – mandatory and optional.

Fig. 1. Mandatory and optional educational models

Year 1, e.g. Year 1, e.g. Anatomy and physiology Anatomy and physiology Science and instrumentation Science and instrumentation Plain %lm radiography Plain %lm radiography EANM Year 2, e.g. Year 2, e.g. Psychology Psychology Adapted x-ray technique Adapted x-ray technique Specialised x-ray procedures Specialised x-ray procedures

PET-CT MR

Ultrasound Nuclear medicine

Figure 1 illustrates two %ctitious models for ra- which they would wish to gain competence diographer training. If the mandatory model is upon quali%cation and the range of choices selected then all students would have to take could include PET-CT, ultrasound, magnetic and also pass the PET-CT training and education resonance and nuclear medicine. In the optional within the third year. If the demand for trained and mandatory models, the decision as to how professionals to work within PET-CT is relatively much time is devoted to the training and educa- low on quali%cation, the optional approach tion will be re+ected in the range of competen- might be preferred. By contrast, the optional cies that are required to be achieved and also in model allows students to select the topic in the depth and level to which the theory is taken.

89 Not all countries include PET-CT %rst post com- raphers or technologists who wish to specialise petencies within formative professional educa- within PET-CT would study either a non-degree tion, one example being the United Kingdom. award bearing quali%cation (‘post-basic’) or a In such cases PET-CT education and training higher degree award bearing quali%cation would be provided within a post-graduate or (‘Post Graduate Diploma’ / ‘Masters of Science post-basic framework. In this case those radiog- degree’). These options are illustrated in Fig. 2.

Fig. 2. Options for education and training in PET-CT

Option A Option B

3-year BSc in diagnostic or therapeutic radiography 3-year BSc in diagnostic or therapeutic radiography

18- to 24-month post-graduate award 18- to 24-month post-graduate 3-month post- (e.g. MSc in nuclear medicine to include award (e.g. MSc) in nuclear basic award in PET-CT theory and practice) medicine) PET-CT © P. Hogg © P.

3-month post-basic award in PET-CT

In option A (Fig. 2), PET-CT theory and practice ing the latter option would also subsequently would be included within the postgraduate take the post-basic award in PET-CT – should nuclear medicine award. Option B would o(er they wish to have specialist skills in PET-CT. Op- the opportunity to study either the 3-month tions A and B currently exist within the UK. The non-degree-bearing post-basic programme choice between these options may depend on in PET-CT or to select a post-graduate pro- what is available locally, the educational back- gramme in nuclear medicine (which might not ground of the radiographer (or technologist) contain PET-CT, unlike option A). Those select- and the clinical role requirements.

90 Chapter 8: Radiographer and technologist competencies – education and training in PET-CT

Obviously, selecting the post-basic or post- technologists. In the context of multi-profes- graduate route into PET-CT would mean that sional education, certain factors require con- a professional identity and knowledge base sideration if such well-designed experiences would have already been established and that are to be achieved: generic matters such as ethics, patient care and management and ‘the sciences’ would t Prior knowledge and skills of potential have already been studied successfully within students the formative professional training and education. The limiting factor in selecting this route t Potential de%ciencies in student knowl- would be the additional total time required to edge and skill that are not covered within train somebody to be competent to practice the PET-CT educational programme and PET-CT – around 5-6 years. Nonetheless, taking the requirement for robust entry require- EANM the lengthier route would add to the profes- ments sional’s skill and knowledge and this is already valued and recognised in the educational op- In some respects the %rst of these factors is portunities a(orded to the medical profes- easier to address than the second. Let us con- sion, whose period of education and training sider a scenario to illustrate the point – a newly can extend to around 10-12 years. Aside from quali%ed radiographer contrasted against a the potential educational pathways already radiographer who quali%ed in 1985. The newly detailed, there are further complexities to quali%ed radiographer is likely to have knowl- consider: edge of and %rst post competence skills in using a CT scanner. This would be required to t Radiographers and technologists who are cope with the job demands of working as a ra- already quali%ed with no PET or CT or PET- diographer in an accident centre at night and CT experience / knowledge during the weekend. By contrast, the radiographer who quali%ed in 1985 and then quickly t Other professionals, such as nurses and moved solely into nuclear medicine may not physicists, who may also wish to gain com- have obtained the CT competencies. If the petence in PET-CT imaging PET-CT programme of study were to cover the fundamentals of CT, including matters like Assuming that the PET-CT education and acquisition parameters, post processing and training experiences are well designed then patient positioning then the newly quali%ed the post-basic and post-graduate models radiographer would likely not bene%t from should meet the needs of many professional that education and time would be wasted. In groups, including quali%ed radiographers and such a case the robust application of Accredi-

91 tation of Prior Learning would allow for knowl- or similar problems is to make clear the entry edge and skill to be valued and accredited to requirements of the programme of study. These that potential student so that they would only entry requirements could be articulated quite attend the required course elements. Obvi- simply by stating nationally recognised quali%- ously, this form of negotiated learning would cations and then Accreditation of Prior Learning become more complex as more professional could be included as a legitimate alternative to groups participated. meet the requirement. For instance, the PET-CT programme entry requirement might be: The second factor is more complex. Let us consider two examples – nurses and physicists. We 1. A recognised quali%cation in Radiography have not yet considered what the programme or Nuclear Medicine Technology or Medi- of study might include, so for the moment cal Physics or Nursing we need to make informed but simplistic as- sumptions to make the examples clearer. Let 2. School-level leaving certi%cate in physics us assume that the PET-CT programme would or Accreditation in Prior Learning cover (a) patient management and care with speci%c reference only to PET-CT and (b) sci- 3. School level certi%cate in human biology ence and technology of PET-CT speci%c only or Accreditation in Prior Learning to PET-CT, without background information on radioactivity. These decisions could be eas- 4. Year 1 nursing skills and knowledge in pa- ily justi%ed in light of the professional groups tient care and management; a recognised most likely to enter the programme of study Nuclear Medicine Technologist quali%ca- (i.e. radiographers and technologists), in that tion; a recognised Radiography quali%ca- they should have already studied and been tion; or Accreditation in Prior Learning examined on generic matters of patient care and management and also background in- To illustrate, a radiographer who quali%ed one formation on radioactivity. If nurse formative year ago would o(er their nationally recog- education does not include the background nised certi%cate or BSc for entry to the PET-CT information on radioactivity that is required programme. This would satisfy point 1. They in order to develop a particular knowledge of could demonstrate, using Accreditation of PET and CT then this will present a problem. Prior Learning, that points 2-4 are also cov- Similarly, if physicist formative education does ered through their nationally recognised BSc not include the requisite aspects of generic pa- or certi%cate by simply copying the learning tient care and management then this, too, will outcomes from that BSc or certi%cate and en- present a problem. The way to overcome both closing them with the application form.

92 Chapter 8: Radiographer and technologist competencies – education and training in PET-CT

In terms of economy of time and money it t Written assignment is essential that the curriculum takes into ac- Individual count what potential students already know Group and can do. This will avoid attendance at class- es in which no new knowledge is acquired. It is worth noting that examinations are good In terms of being safe to practice, the entry for testing a wide breadth of knowledge but requirements and assessment methods need they are quite poor at assessing depth of to be robust to ensure that the clinical and understanding and also application of that theoretical learning outcomes are met. knowledge. Written assignments are good at testing depth of understanding and applica- What will be assessed in the programme tion, but they are poor at testing breadth of of study? knowledge. Clearly an appropriate blend of EANM Competence to practice must be assured on assignment types needs to be considered for successful completion of a programme of assessing theory. study and to achieve this, clinical competency must be tested and the theory on which such Clinical practice can be assessed in di(erent competence is based must also be tested. Vari- ways, too, for example: ous approaches to testing theory and practice and their integration exist and an assessment t Objective structured clinical examination strategy should be developed to ensure that the student is safe and %t to practice. t Clinical assessment (Performing clinical practice whilst being observed and Theory can be assessed in di(erent ways, for ‘scored’) example: t Portfolio and case study compilation t Examination Seen The integration of theory and practice can Unseen also be assessed in di(erent ways, for example: Open book Multiple choice questionnaire t Re+ective reports t Objective structured examination t Portfolios and case study compilation t Viva oral examination t Objective structured clinical examination

93 Each of the above has its value and limita- ance procedure could be through use of an tions, and a well-designed programme of external examiner system and delivery of the study would contain an appropriate range programme of study by an organisation that and balance of assessment methods within permits external educational audit and pub- its assessment strategy. lication of the results ready for public access. An example of the latter would be a univer- What should be learnt and what sity. The ultimate aim of external accreditation competencies should be acquired? and audit is to ensure that the programme of There will be variation between countries and study and therefore the students are %t for individuals in response to this question. For purpose so as to protect patients from poor instance, in some countries only medically clinical practice. quali%ed sta( may administer the PET radiopharmaceutical, whilst in others appropriately It is likely that a PET-CT syllabus would attract trained, quali%ed, competent and insured per- signi%cant debate, in terms of what should be sons can do this task – clearly this will include included. Some are likely to argue that ‘facts medically quali%ed sta(. However, for radiog- and topics’ are essential and that the student raphers and technologists there would be a should rote learn a broad range of information. core set of competencies and principles that Such a list would include minute detail and be would be fairly well recognised internationally, very wide in terms of the topics covered. The and the %nal element of this chapter seeks to alternative approach would be to consider consider those aspects. what knowledge is required to be competent and to what level that knowledge should be The %rst important principle is that the pro- taken. This would require learning outcomes gramme of PET-CT study should have external to be written and considerable thought would accreditation and be open for public and pro- be required to link syllabus detail to those out- fessional audit and accountability. This would comes. This is a more thorough approach than involve at least one professional body approv- simply listing syllabus content. Examples of ing the curriculum prior to it being delivered. topics that could be included are indicated If appropriate, a regulatory (legal) body might below: also need to accredit it. Public and professional scrutiny would come through an external t Minimisation of dose to patients (PET and CT) quality inspection mechanism. Internal self- regulation would be discouraged as standards t Radiation protection of sta( could not be assured or veri%ed. Methods of achieving a robust educational quality assur- t Maximisation of image quality

94 Chapter 8: Radiographer and technologist competencies – education and training in PET-CT

t Care of patient A comprehensive syllabus is provided by So- ciety of Nuclear Medicine/American Society t PET tracer production of Radiologic Technologists [1]. In light of the arguments already set out within this chapter, t PET chemistry with respect to radiophar- it would be advisable to consider prior knowl- maceuticals edge and skill, which could call into question the need to ‘teach’ that curriculum. In contrast t PET instrumentation – construction and to the American approach is that o(ered by principles of operation the Society and College of Radiographers [2] (UK). In their document they pay attention to t Issues associated with PET and CT as a hy- competencies required by radiographers and brid unit, including registration technologists while leaving the detail of the EANM syllabus to educational providers who would t Quality control and assurance ‘of the whole work in collaboration with clinical PET-CT context’ centres. Table 1 [2] illustrates examples of the clinical competencies required for entry level t Diagnostic procedures PET-CT practice within the UK for radiographers and technologists (note that this is level t Procedures for therapy planning 2 of a four-level structure; level 1 is assistant practitioner). t PET tracers and their administration t Non-radioactive medicines/drugs within the diagnostic procedures and therapy planning t Computer processing

95 Table 1: Clinical competencies required by practitioners (level 2 of four levels) for PET-CT

Practitioners need to possess a current Practitioners’ level of knowledge should be knowledge and understanding of: su$cient to enable them to:

t the risk-bene%t philosophy as applied to nuclear t identify and respond to those situations that are medicine and hybrid imaging beyond the scope of practice of the assistant practitioner t the scienti%c and legal basis for nuclear medicine and hybrid imaging examinations and inter- t select, plan, implement, manage and evaluate ventions, including the legal basis and practical imaging procedures that are appropriate to, and implementation of radiation protection laws take account of, individuals’ health status, environment and needs and the legal framework t the legal basis of supply, administration and pre- of practice scribing of medicines t participate e(ectively within multi-professional t drug interactions, pharmacology and adverse re- health care and multi-agency teams and in actions of drugs commonly encountered within health care environments both within and be- imaging settings, with a particular emphasis on yond clinical imaging services radiopharmaceuticals and contrast agents t analyse systematically, evaluate and act upon all t the methods of administration of drugs, includ- data and information relevant to the care and ing the associated health, safety and legal issues management of the patient t developments and trends in the science and t be able to acquire and process CT images and practice of nuclear medicine data that have clinical relevance within nuclear t the safe practice of CT when used as an adjunct medicine, observing the principles of exposure to a nuclear medicine service (i.e. PET-CT) optimisation particularly with respect to attenu- t the principles underpinning moving and han- ation correction and diagnostic CT dling, the principles underpinning emergency aid t assess patients’ needs and, where necessary, refer and the principles (including health, safety and to other relevant health care professionals legal considerations) underpinning assessment t be able to manipulate written and image data t monitoring and care of the patient before, during in di(ering formats for the bene%t of the patient and after examination t o(er the highest standards of care in both physical and psychological respects in all aspects of nuclear medicine and hybrid imaging examinations and interventions in order to ensure e(ective procedures t make informed, sensitive and ethically sound professional judgements in relation to imaging procedures in which they are involved t ensure that consent given by patients to procedures is ‘informed’ t apply safe and e(ective moving and handling skills in order to protect all individuals

96 Chapter 8: Radiographer and technologist competencies – education and training in PET-CT

Worthy of note are the competencies required amples of which are shown in Table 2 [2]. Please for the highest grade of radiographer and tech- note that there is one grade between Tables 1 nologist (consultant) within the UK for PET-CT, ex- and 2; this grade is called “advanced practitioner”.

Table 2: Clinical competencies required by consultants (level 4 of four levels) for PET-CT

Consultants’ depth and breadth of knowledge and expertise in nuclear medicine practice and hybrid imaging will enable them to:

t ide ntify and respond to those situations that are beyond the scope of practice of the advanced practitioner, providing training, supervision and mentorship as part of the role t e(ectively lead the clinical team in the delivery of the nuclear medicine service, including hybrid imaging t exhibit expert clinical practice in managing complete episodes of care that lead to satisfactory patient EANM outcomes and/or health gains, including determining the suitability of clinical requests t deliver a whole-system, patient-focussed, approach rooted in a multi-professional perspective lead and/or represent the team at multidisciplinary meetings t provide clinical leadership locally and across professional/organisational boundaries at a national and/or international level where appropriate t manage personal case loads, including wide-ranging decision making and the provision of a clinical report t engage in the development and advancement of innovative practice by means of active involvement in research t be accountable for safety, legal and clinical governance issues for nuclear medicine and hybrid imaging practice t evaluate, identify gaps in and integrate the research evidence base into practice such that expert professional judgements can be exercised routinely t supply and administer medicines wit hin the legal framework

The coverage of material by the syllabus needs signed, the depth versus breadth factor needs to be balanced against the level to which the signi%cant consideration and this will be driven material is learnt and subsequently applied. by the level to which the radiographer or tech- As a rule of thumb, for the same programme nologist will operate on quali%cation. Again, for length, the broader the syllabus coverage, the similar programme lengths, a rule of thumb more super%cial is the material learnt and ap- is that a broad syllabus range means a lower plied. Conversely, the narrower the syllabus level of practice compared with a narrow syl- coverage, the greater can be the depth and labus range taken to a much greater depth. This application. When the programme is being de- trade o( should be determined when a ratio-

97 References Chapter 8

References nale for a PET-CT programme is proposed, and 1. Society of Nuclear Medicine and American Society it is likely that this will result in di(erences be- of Radiologic Technologists. Positron emission tomo- tween countries. Such di(erences are likely to graphy (PET)-computed tomography (CT) curriculum. https://www.asrt.org/media/Pdf/PETCTCurriculumAc- be driven by political, legal and clinical factors. cepted021704.pdf . 2004

2. Society and College of Radiographers. Learning and Summary development framework for hybrid nuclear medicine/ Several philosophical educational debates computed tomography practice (SPECT-CT/PET-CT). need to be concluded prior to deciding on www.sor.org . 2009 the %ne detail of a PET-CT curriculum and syl- Suggested Reading labus. The broader curricular considerations Iwasiw CL. Curriculum development in nursing education. 2nd ed. Sudbury, Mass.: Jones and Bartlett; 2008. encompass matters such as where PET-CT will be learnt and applied within the profession- Nyatanga L, Forman D, Fox J. Good practice in the accreditation of prior learning. London: Continuum International al’s overall educational experience. Country- Publishing Group (Cassell Education); 1998. speci%c issues will require due consideration, Anema M, McCoy J. Competency based nursing education: and these will inform the level of responsibil- guide to achieving outstanding learner outcomes. Berlin ity that could be held by radiographers and Heidelberg New York: Springer; 2009. technologists. Such debates will drive the syllabus range and depth to engender the competencies that are to the required level for practice. Given the multi-professional nature of nuclear medicine and PET-CT, there is a need to design curricula that meet a range of professional needs. This means that an adap- tive curriculum will be required that takes into account and also values prior knowledge and skill; but whatever the route that is taken, the educational outcomes must be assured. Per- haps the %nal and most important part of the educational provision concerns the need for public accountability through an open and robust quality assurance system.

98 Layout andDesign: asperdate ofpreparation:Information August 2010 for ofthisinformation. istaken No responsibility thecorrectness This copy isareprint ofthe2010version. Content: URL: www.eanm.org E-mail: [email protected] Tel: +43-(0)1-2128030,Fax: +43-(0)1-21280309 Hollandstrasse 14,1020 Vienna, Austria EANM Technologist Committee European Association ofNuclearMedicine Publisher: ISSN: 2079-3138 ISBN: 978-3-902785-00-8 Imprint URL: www.laberdruck.at E-mail: [email protected] Tel: +43-(0)6272-7135-426,Fax +43-(0)6272-7135-499 Austria 46,5110Oberndorf, Michael-Rottmayr-Straße Laber-Druck Ges.m.b.H. Printing: URL: www.kreativ-sacher.at E-mail: o#[email protected] Tel: +43-(0)1/4165227,Fax: +43-(0)1/4168526 Linzer Straße358a/1/7,1140 Vienna, Austria ·Mag. Evelynekreativ Sacher 99

EANM

Principles and Practice of PET/CT Part 1 A Technologist’s Guide