OPTIMISATION AND ESTABLISHMENT OF DIAGNOSTIC ......To Professor Eliseo Vaño, Cathedratic Professor of Medical Physics of the Complutense University of Madrid, a world-renowned expert

1

Graciano do Nascimento Nobre Paulo

OPTIMISATION AND ESTABLISHMENT OF

DIAGNOSTIC REFERENCE LEVELS

IN PAEDIATRIC PLAIN RADIOGRAPHY

Tese de Doutoramento em Ciências da Saúde - Ramo das Tecnologias da Saúde, orientada pelo Senhor Professor Doutor Eliseo Vaño e pelo Senhor Professor Doutor Adriano Rodrigues e

apresentada à Faculdade de Medicina da Universidade de Coimbra.

Setembro de 2015

Optimisation and establishment of Diagnostic Reference

Levels in paediatric plain radiography

Graciano do Nascimento Nobre Paulo

Tese de Doutoramento em Ciências da Saúde-Ramo das Tecnologias da Saúde apresentada à Faculdade de Medicina da Universidade de Coimbra

Orientadores

Professor Doutor Eliseo Vaño, Professor Catedrático da Faculdade de Medicina da Universidade Complutense de Madrid

Professor Doutor Adriano Rodrigues, Professor Associado, Faculdade de Medicina da Universidade de Coimbra

September de 2015

OptimisationandestablishmentofDiagnosticReferenceLevelsinpaediatricplainradiography

GracianodoNascimentoNobrePaulo

II

§



III

AcknowledgementsThedevelopmentof this thesiswouldnothavebeenpossiblewithout thehelpofseveral friends that followedme from the firstminute.Toallof themmysincerethankyou,forsharingoutstandingmomentsandhelpingmetoremovethestonesfromalongandwindingroad.

To Professor Eliseo Vaño, Cathedratic Professor of Medical Physics of theComplutense University of Madrid, a world-renowned expert in the field ofRadiationProtectiona special thanks, forhavingaccepted to supervise this thesisandforallthatIhavelearnedfromhiminthepastyears.Ihavenowordstoexpressmy gratitude for his permanent support and outstanding advices. His vision andknowledge has been the main contributor for building bridges between healthprofessionals towards a continuous improvement in the quality and safety ofhealthcareservicesallaroundtheworld.

To Professor Adriano Rodrigues, a distinguished Doctor of Internal Medicine ofCoimbra Hospital and University Centre, Professor of theMedical Faculty of theUniversityofCoimbra,thathasdedicatedhislifetothedevelopmentoftheNuclearSciencesApplied toHealth inPortugal,aspecial thanks for trusting inmeand forhaving accepted to supervise this thesis. Thank you for encouragingme tomoveforwardandformakingmebelievethatitwaspossible.

To Professor Joana Santos, Director of the Medical Imaging and RadiotherapyDepartment of ESTESC-Coimbra Health School, my recognition and gratitude.Withouthertheachievementof thisobjectivewouldnothavebeenpossible.Hersupport,energyandengagementwerethemainpillarsforthedevelopmentofthisthesis.

To Professor Filipe CaseiroAlves, a distinguishedMedical Radiologist, CathedraticProfessorof theMedical Facultyof theUniversityofCoimbraandDirectorof theRadiology Department of the Coimbra Hospital and University Centre, a specialthank you forhispermanent support and foropeningdoors forwhathasbeenafruitfulcooperationforappliedresearchinthefieldofmedicalimaging.

To Dr. Amélia Estevão, a distinguished Medical Radiologist of the RadiologyDepartmentoftheCoimbraHospitalandUniversityCentre,aspecialthanksforheroutstandingcontributionforobtainingtheapprovaltodevelopthisthesis.

To all Radiologists and Radiographers from the three Portuguese PaediatricHospitals,especiallytheSeniorRadiographers,FilomenaOliveira,FernandaAndré,AldaPinto,CristinaAlmeidaandDalilaFerreira,aspecialthanksfortheassistanceprovidedduringthedatacollection.

ToalltheRadiologistsfromthePaediatricHospitalofCoimbraaspecialthanksfortheircontributiontothisstudy.



IV

ToDr. PintoMachado, a seniorMedical Radiologist a special thanks for his help,advice and cooperation and for all that he has taught me through out myprofessionallife.

TomyProfessor,JoãoJoséPedrosodeLima,oneofthemostprominentPortugueseprofessorofmedicalphysics,myinspirationandreference,bothasapersonandasanacademic,myspecialthanks.

To all Professors from the Medical Imaging and Radiotherapy Department ofESTESC-CoimbraHealthSchoolaspecialthanksfortheirsupport.

To all Professors from ESTESC-Coimbra Health School a special thanks for theirsupportandunderstanding.

TomycolleaguesfromtheBoardofManagementofESTESC-CoimbraHealthSchool,ProfessorJorgeCondeandProfessorAnaFerreiraaspecialthanksfortheirsupportandpatience.

ToallmyRadiographyandMedicalImaging&RadiotherapystudentswithwhomIhave the privilege to learn everyday, a special thanks for being the realambassadorsofESTESC-CoimbraHealthSchool.

ToallthestaffofESTESC-CoimbraHealthSchoolwithwhomIhavetheprivilegetoworkwithaspecialthanks.

Iwould liketodedicatethisthesistomyfamilystartingbymyparentsCarlosandAliceforbeingmyheroesandforteachingmethefundamentalvaluesoflife.

TomyyoungersisterCarlaforalwaysbeingthereformewhenIneeded.

Tomytwosons,CésarandInês,theessentialpartofmylife,towhomIapologiseforallowingmyworktotakethetimethatIshouldhavededicatedtothem.

To Laila, my wife, an outstanding woman. I can’t find enough words in thedictionary to describe what you represent in my life. Nothing would have beenpossiblewithout you. This thesis is yours. Iwill neverbeable to compensate thetimeoutofhomeandwhenathomeclosedinmyoffice.Thankyouforsharingyourlifewithme.



V

Musicismypassionanditwasmycompanionduringdaysandnightswhilewritingthisthesis.IwouldliketoquoteoneofthemostpopularandinfluentialmusiciansofthehistoryofRock&Roll:

No,youcan'talwaysgetwhatyouwantButifyoutrysometime,youjustmightfind

Yougetwhatyouneed

SirMichaelPhilip"Mick"Jagger



VI

§



VII

AbstractPurpose: This study aimed to propose Diagnostic Reference Levels (DRLs) inpaediatric plain radiography and to optimise the most frequent paediatric plainradiography examinations in Portugal following an analysis and evaluation ofcurrentpractice.

Methodsandmaterials:Anthropometricdata(weight,patientheightandthicknessof the irradiated anatomy) was collected from 9,935 patients referred for aradiography procedure to one of the three dedicated paediatric hospitals inPortugal. National DRLs were calculated for the three most frequent X-rayproceduresatthethreehospitals:chestAP/PAprojection;abdomenAPprojection;pelvis AP projection. Exposure factors and patient dose were collectedprospectively at the clinical sites. In order to analyse the relationship betweenexposure factors, theuseof technical featuresanddose,experimental testsweremadeusingtwoanthropomorphicphantoms:a)CIRSTMATOMmodel705®;height:110cm,weight:19kgandb)KyotokagakuTMmodelPBU-60®;height:165cm,weight:50kg.Afterphantomdatacollection,anobjectiveimageanalysiswasperformedbyanalysingthevariationofthemeanvalueofthestandarddeviation,measuredwithOsiriX® software (Pixmeo, Switzerland). After proposing new exposure criteria, aVisual Grading Characteristic image quality evaluation was performed blindly byfour paediatric radiologists, each with a minimum of 10 years of professionalexperience,usinganatomicalcriteriascoring.

Results:AhighheterogeneityofpracticewasfoundandtheestablishedPortugueseDRLvalues(KermaAirProductpercentile75,KAPP75andEntranceSurfaceAirkermapercentile 75, ESAKP75) were higher than the most recent published data. ThenationalDRLsestablishedforPortugalare:CHEST:KAPP75,13mGy.cm2,19mGy.cm2,60mGy.cm2,134mGy.cm2,94mGy.cm2,respectivelyforagegroups<1,1-<5,5-<10,10-<16, 16-≤18. ABDOMEN: KAPP75, 25mGy.cm2, 84mGy.cm2, 140mGy.cm2,442mGy.cm2, 1401mGy.cm2, respectively for age groups <1, 1-<5, 5-<10, 10-<16,16-≤18. PELVIS: KAPP75, 29mGy.cm2, 75mGy.cm2, 143mGy.cm2, 585mGy.cm2,839mGy.cm2,respectivelyforagegroups<1,1-<5,5-<10,10-<16,16-≤18.

DRLs by patientweight groupshavebeenestablished for the first time. Thepostoptimisation DRLs by patient weight groups are: CHEST: KAPP75, 9mGy.cm2,10mGy.cm2, 15mGy.cm2, 32mGy.cm2, 57mGy.cm2, respectively for weight groups<5kg; 5-<15kg; 15-<30kg; 30-<50kg; ≥50kg. ABDOMEN: KAPP75, 10mGy.cm2,20mGy.cm2,61mGy.cm2,203mGy.cm2,225mGy.cm2,respectivelyforweightgroups<5kg;5-<15kg;15-<30kg;30-<50kg;≥50kg.PELVIS:KAPP75,15mGy.cm2,18mGy.cm2,45mGy.cm2,75mGy.cm2,79mGy.cm2,respectivelyforweightgroups<5kg;5-<15kg;15-<30kg;30-<50kg;≥50kg.



VIII

ESAKP75DRLs for bothpatient age andweight groupswere alsoobtained and aredescribedinthethesis.

Significant dose reduction was achieved through the implementation of anoptimisation programme: an average reduction of 41% and 18% on KAPP75 andESAKP75,respectivelyforchestplainradiography;anaveragereductionof58%and53% on KAPP75 and ESAKP75, respectively for abdomen plain radiography; and anaverage reduction of 47% and 48%on KAPP75 and ESAKP75, respectively for pelvisplainradiography.

Conclusion: Portuguese DRLs for plain radiography were obtained for paediatricplainradiography(chestAP/PA,abdomenandpelvis).Experimentalphantomtestsidentifiedadequateplainradiographyexposurecriteria,validatedbyobjectiveandsubjectiveimagequalityanalysis.Thenewexposurecriteriawereputintopracticeinoneofthepaediatrichospitals,by introducinganoptimisationprogramme.Theimplementation of the optimisation programme allowed a significant dosereductiontopaediatricpatients,withoutcompromisingimagequality.

Keywords:diagnosticreferencelevels;paediatricradiology;radiationprotection;optimisation.



IX

ResumoObjetivo: Este estudo teve como objetivo propor Níveis de Referência deDiagnóstico (NRD) para a radiologia convencional pediátrica e otimizar osprocedimentosradiológicosmaisfrequentesemPortugal,partindodeumaanáliseeavaliaçãodaspráticasatuais.

Materiais e Métodos: Foram recolhidos dados antropométricos (peso, altura eespessura anatómica da estrutura radiografada) de 9.935 doentes, referenciadosparaumexame radiológico, paraumdos trêshospitais pediátricos existentes emPortugal. Os NRDs nacionais foram calculados para os três procedimentosradiológicosmais frequentes: radiografiadotóraxAP/PA;radiografiadoabdómenAP;radiografiadabaciaAP.Osfactoresdeexposiçãoassociadosaosprocedimentosbemcomoosvaloresdedosenodoenteforamrecolhidosdeformaprospectivaemcadaumdoshospitais.

Por forma a analisar a relação entre os parâmetros de exposição e a respectivadose,foiefetuadoumestudoexperimentalusandodoisfantomasantropomórficos:a)modelo CIRSTM ATOM 705®; altura: 110 centímetros, peso: 19 kg e b)modeloKyoto kagakuTM PBU-60®; altura: 165 centímetros, peso: 50 kg. Na sequência doestudo experimental nos fantomas, foi efetuada uma avaliação objectiva dasimagens,atravésdaanálisedavariaçãodovalormédiododesvio-padrão,medidoscom o software OsiriX® (Pixmeo, Suíça). Com base nos resultados obtidos forampropostos novos parâmetros de exposição, para cada um dos procedimentos emestudo.Paravalidarosnovosparâmetrosdeexposiçãoemprocedimentosclínicosfoiefetuadaumaavaliaçãosubjetivadaqualidadedasimagensradiológicas,atravésdométodo Visual Grading Charateristics (VGC), realizada de forma independenteporquatroespecialistasemradiologiapediátrica,cadaumcomummínimode10anos de experiência profissional utilizando, para tal, critérios de avaliaçãoanatômica.

Resultados: Foi identificadaumagrandeheterogeneidadena formadeefetuarosprocedimentosradiológicosemestudo,tendosidocalculadososNRDparaPortugal,definidoscomopercentil75doProdutoDose-Área,(KAPP75)epercentil75dadoseá entradada pele (ESAKP75) que se revelarammais elevados quando comparadoscomosdadosmais recentespublicadosna literatura.OsNRDsestabelecidosparaPortugal são: TÓRAX AP/PA: KAPP75, 13mGy.cm2, 19mGy.cm2, 60mGy.cm2,134mGy.cm2,94mGy.cm2,respectivamenteparaosgruposetários<1,1-<5,5-<10,10-<16, 16-≤18. ABDOMEN: KAPP75, 25mGy.cm2, 84mGy.cm2, 140mGy.cm2,442mGy.cm2, 1401mGy.cm2, respectivamenteparaos gruposetários<1, 1-<5, 5-<10, 10-<16, 16≤18. BACIA: KAPP75, 29mGy.cm2, 75mGy.cm2, 143mGy.cm2,



X

585mGy.cm2, 839mGy.cm2, respectivamente para os grupos etários <1, 1- <5, 5-<10,10-<16,16-≤18.

Foram tambémestabelecidos pela primeira vez osNRDs por grupos de peso dosdoentes.OsNRDsobtidosapósoprocessodeotimizaçãoporgruposdepesodosdoentes são: TORAX AP/PA: KAPP75, 9mGy.cm2, 10mGy.cm2, 15mGy.cm2,32mGy.cm2, 57mGy.cm2, respectivamente para os grupos de peso <5kg; 5-<15kg;15-<30kg; 30- <50kg; ≥50kg. ABDÓMEN: KAPP75, 10mGy.cm2, 20mGy.cm2,61mGy.cm2,203mGy.cm2,225mGy.cm2, respectivamenteparaosgruposdepeso,<5kg;5-<15kg;15-<30kg;30-<50kg;≥50kg.BACIA:KAPP75,15mGy.cm2,18mGy.cm2,45mGy.cm2,75mGy.cm2,79mGy.cm2, respectivamenteparaosgruposdepeso<5kg;5-<15kg;15-<30kg;30-<50kg;≥50kg.

OsNRDsrelativosàESAKP75paraambososgruposdeidadeedepesodosdoentestambémforamobtidaseestãodescritosnatese.

Foi conseguida uma redução significativa na dose nos doentes após aimplementação do programa de otimização: uma redução média de 41% e 18%respectivamente nos valores de KAPP75 e de ESAKP75 para a radiografia do tóraxAP/PA;umareduçãomédiade58%e53%respectivamentenosvaloresdeKAPP75ede ESAKP75, para a radiografia do abdómen; uma redução média de 47% e 48%respectivamentenosvaloresdeKAPP75edeESAKP75,paraaradiografiadabacia.

Conclusão:ForamdefinidososNRDsnacionaisparaasradiografiasdoTóraxAP/PA,Abdómen e Bacia. O estudo experimental efetuado permitiu definir critérios deexposiçãomaisadequadosedevidamentevalidadosatravésdaavaliaçãoobjectivaesubjetivadasimagensradiológicas.Aimplementaçãodoprogramadeotimizaçãopermitiu uma significativa redução da dose nos doentes pediátricos semcomprometeraqualidadedaimagem.

Palavraschave:níveisdereferênciadediagnóstico;radiologiapediátrica;proteçãoradiológica;otimização.



11

TableofContentsFIGUREINDEX.................................................................................................................13

TABLEINDEX...................................................................................................................15

EQUATIONINDEX...........................................................................................................19

ABBREVIATIONINDEX....................................................................................................21

INTRODUCTIONANDOBJECTIVES...................................................................................25

1 BACKGROUND..........................................................................................................291.1 PORTUGUESEHEALTHCARECONTEXT...............................................................................311.2 EUROPEANANDPORTUGUESELEGALFRAMEWORKSONIONISINGRADIATION..........................331.3 THESHIFTOFPARADIGMINMEDICALIMAGING..................................................................371.4 PLAINRADIOGRAPHYDETECTORSYSTEMS........................................................................39

1.4.1 Screen-filmSystems..........................................................................................391.4.2 DigitalSystems..................................................................................................43

1.5 DOSEDESCRIPTORSINRADIOGRAPHY...............................................................................511.6 RISKSINPAEDIATRICIMAGING........................................................................................571.7 THEINTERNATIONALCONTEXTOFDIAGNOSTICREFERENCELEVELS........................................611.8 THEPORTUGUESECONTEXTOFDIAGNOSTICREFERENCELEVELS...........................................67

2 ESTABLISHMENTOFDRLSINPAEDIATRICPLAINRADIOGRAPHY...............................712.1 MATERIALSANDMETHODSTODETERMINENATIONALDRLSFORCHEST,ABDOMENANDPELVIS

PLAINRADIOGRAPHY...............................................................................................................732.2 RESULTSOFNATIONALDRLSFORCHEST,ABDOMENANDPELVISPLAINRADIOGRAPHY..............75

2.2.1 NationalDRLsbyagegroups............................................................................812.2.2 NationalDRLsbyweightgroups.......................................................................832.2.3 NationalversuslocalDRLs................................................................................85

2.3 LIMITATIONSOFSECTION2............................................................................................91

3 PLAINRADIOGRAPHYOPTIMISATIONPHANTOMTESTS...........................................933.1 OPTIMISATIONINPLAINRADIOGRAPHY.............................................................................933.2 EXPERIMENTALTESTSWITHANTHROPOMORPHICPHANTOMS(OBJECTIVEIMAGEANALYSIS)......97

3.2.1 Methodologyofexperimentaltestswithanthropomorphicphantoms............973.2.2 Resultsofphantomsexperimentaltests...........................................................99

3.3 OPTIMISEDEXPOSURECRITERIAFORCHEST,ABDOMENANDPELVISPLAINRADIOGRAPHY........1053.4 SUBJECTIVEANALYSISOFIMAGEQUALITY(METHODOLOGYANDRESULTS).............................1093.5 ASSESSINGTHEUSEOFELECTRONICCROPPINGINPLAINIMAGING.......................................1153.6 LIMITATIONSOFSECTION3..........................................................................................117

4 IMPACTOFTHEOPTIMISATIONPROGRAMMEONPATIENTDOSES........................1194.1 MATERIALANDMETHODSTOASSESSTHEIMPACTOFOPTIMISATIONONPATIENTDOSES.........1194.2 RESULTSOFTHEIMPACTOFOPTIMISATIONONPATIENTDOSES...........................................121

5 POSTOPTIMISATIONDRLS......................................................................................1295.1 NEWDRLSBYAGEGROUP...........................................................................................129



12

5.2 NEWDRLSBYWEIGHTGROUP......................................................................................131

6 DISCUSSION............................................................................................................1336.1 ABOUTPATIENTCHARACTERISTICS.................................................................................1336.2 ABOUTEXPOSUREPARAMETERSOFPHASE1....................................................................1336.3 ABOUTNATIONALDRLS...............................................................................................1356.4 ABOUTTHEOPTIMISATIONTESTS...................................................................................1396.5 ABOUTTHEIMPACTOFTHEOPTIMISATIONPROGRAMMEONPATIENTDOSE..........................141

CONCLUSIONS..............................................................................................................143

REFERENCES.................................................................................................................147



13

FigureIndexFigure1:Schematicmapofresearchactivityandphasesoftheoverallthesis........27

Figure2:PortuguesemapindicatingtheRegionalHealthAuthorities(RHA)...........32

Figure3:Screen-filmreceptor...................................................................................40

Figure4:AHurtherandDriffieldcurve.....................................................................41

Figure5:Taxonomyforplainradiographydigitalsystems........................................43

Figure6:SchematicrepresentationofaCRreadersystem......................................44

Figure7:SchematicrepresentationofDRsystems...................................................46

Figure8:DynamicrangeindigitalandS/Fsystems..................................................48

Figure 9: Schematic representation of a radiograph with some dosimetric andgeometricquantitiesfordeterminationofpatientdose..................................52

Figure10:DRLsforpaediatricplainradiographyinEuropeancountries..................64

Figure11:Patientdistributionbygender..................................................................75

Figure12:Weightperagegroupboxplot..................................................................76

Figure13:Heightperagegroupboxplot...................................................................76

Figure14:BMI(kg/m2)peragegroupboxplot..........................................................77

Figure15:ComparisonoftheHospitals’KAPP75valuewiththe“1stNationalDRL”forchestplainradiography.....................................................................................85

Figure16:Comparisonof theHospitals’ ESAKP75 valuewith the “1stNationalDRL”forchestplainradiography...............................................................................86

Figure17:ComparisonoftheHospitals’KAPP75valuewiththe“1stNationalDRL”forabdomenplainradiography..............................................................................86

Figure18:Comparisonof theHospitals’ ESAKP75 valuewith the “1stNationalDRL”forabdomenplainradiography.........................................................................87

Figure19:ComparisonoftheHospitals’KAPP75valuewiththe“1stNationalDRL”forpelvisplainradiography....................................................................................87

Figure20:ComparisonoftheHospitalESAKP75valuewiththe“1stNationalDRL”forpelvisplainradiography....................................................................................88

Figure21:MeankVvaluesusedbyeachradiographerforchestplainradiographyineachpatientagegroup......................................................................................88

Figure22:Optimisationofclinicalprotocolsforpaediatricimaging.........................95



14

Figure23:Anthropomorphicphantomsusedinexperimentaltests........................97

Figure24:ExampleofROIlocations,foranalyseswithOsiriX®software(AtoE)....98

Figure 25: A: Chest plain radiographywithAEC + central chamber; B: Chest plainradiographywithAEC+lateralrightchamber................................................104

Figure26:ChestVGCanalysisperagegroup..........................................................110

Figure27:AbdomenVGCanalysisperagegroup...................................................111

Figure28:PelvisVGCanalysisperagegroup..........................................................111

Figure29:ExposureTime(ms)values forchestplainradiography:phase1vspostoptimisation....................................................................................................121

Figure 30: Exposure Time (ms) values for abdomenplain radiography: phase 1vspostoptimisation............................................................................................121

Figure31:ExposureTime(ms)valuesforpelvisplainradiography:phase1vspostoptimisation....................................................................................................122



15

TableIndexTable1:Manufacturerexposureindexnameandindicatorofdigitalsystems........54

Table2: ProposedPortugueseCTDRLs for adultMSCTexaminationsdescribedasCTDIvolandDLPvalues.......................................................................................67

Table3:Proposedage-categorisednationalpaediatricCTDRLsdescribedasCTDIvolandDLPvalues...................................................................................................67

Table 4: KAPP75 values (Gy.cm2) for diagnostic paediatric interventional cardiologyprocedures.........................................................................................................69

Table5:PaediatricpatientsweightheightandBMI(byagegroups)........................75

Table6:Chest,abdomenandpelvisthicknessperagegroup...................................77

Table7:ExposureparametersofchestAP/PAprojection.........................................78

Table8:ExposureparametersofabdomenAPprojection........................................79

Table9:ExposureparametersofpelvisAPprojection..............................................80

Table10:KAP&ESAKvaluesforchestAP/PA(byagegroups).................................81

Table11:KAP&ESAKvaluesforabdomenAP(byagegroups)................................81

Table12:KAP&ESAKvaluesforpelvisAP(byagegroups)......................................82

Table13:KAP&ESAKvaluesforchestAP/PA(byweightgroups)............................83

Table14:KAP&ESAKvaluesforabdomenAP(byweightgroups)...........................84

Table15:KAP&ESAKvaluesforpelvisAP(byweightgroups).................................84

Table16:ExperimentaltestsforchestexaminationusingCIRSTMATOMmodel705®...........................................................................................................................99

Table17:Experimental tests forabdomenexaminationusingCIRSTMATOMmodel705®.................................................................................................................100

Table18:ExperimentaltestsforpelvisexaminationusingCIRSTMATOMmodel705®.........................................................................................................................101

Table 19: Experimental tests for chest examination using Kyoto kagakuTM modelPBU-60.............................................................................................................102

Table20:ExperimentaltestsforabdomenexaminationusingKyotokagakuTMmodelPBU-60.............................................................................................................103

Table21:Newexposurecriteriaforchestplainradiography.................................105

Table22:Newexposurecriteriaforabdomenplainradiography...........................106

Table23:Newexposurecriteriaforpelvisplainradiography.................................107



16

Table24: Imageanalysesusinganatomicalcriteriascoringandthefivepointscale.........................................................................................................................109

Table25:VGCanalysisbyanatomicalcriterion......................................................112

Table26:irradiatedversuspostprocessedimagearea..........................................115

Table 27: KAPP75, ESAKP75 and P75 variation values for chest plain radiography:phase1vspostoptimisation(agegroups)......................................................123

Table28:KAPP75,ESAKP75andP75variationvaluesforabdomenplainradiography:phase1vspostoptimisation(agegroups)......................................................124

Table 29: KAPP75, ESAKP75 and P75 variation values for pelvis plain radiography:phase1vspostoptimisation(agegroups)......................................................124

Table 30: KAPP75, ESAKP75 and P75 variation values for chest plain radiography:phase1vspostoptimisation(weightgroups)................................................125

Table31:KAPP75,ESAKP75andP75variationvaluesforabdomenplainradiography:phase1vspostoptimisation(weightgroups)................................................126

Table 32: KAPP75, ESAKP75 and P75 variation values for pelvis plain radiography:phase1vspostoptimisation(weightgroups)................................................127

Table33:NewKAP&ESAKvaluesforchestAP/PA(byagegroups).......................129

Table34:NewKAP&ESAKvaluesforabdomenAP(byagegroups)......................129

Table35:NewKAP&ESAKvaluesforpelvisAP(byagegroups)............................130

Table36:NewKAP&ESAKvaluesforchestAP/PA(byweightgroups).................131

Table37:NewKAP&ESAKvaluesforabdomenAP(byweightgroups)................131

Table38:NewKAP&ESAKvaluesforpelvisAP(byweightgroups).......................132

Table39:ConversionfactorsforKAPunits.............................................................135

Table 40: Comparison of values for chest AP/PA plain radiography ESAKP75 (µGy)withotherpublisheddata...............................................................................136

Table41:ComparisonofvaluesforabdomenplainradiographyESAKP75(µGy)withotherpublisheddata.......................................................................................137

Table 42: Comparison of values for pelvis plain radiography ESAKP75 (µGy) withotherpublisheddata.......................................................................................137

Table43:ComparisonofvaluesforchestplainradiographyKAPP75(mGy.cm2)withotherpublisheddata.......................................................................................137

Table 44: Comparisonof values for abdomenplain radiographyKAPP75 (mGy.cm2)withotherpublisheddata...............................................................................138



17

Table45:ComparisonofvaluesforpelvisplainradiographyKAPP75(mGy.cm2)withotherpublisheddata.......................................................................................138



18

§



19

EquationIndexEquation1:DetectiveQuantumEfficiency................................................................46

Equation2:EquivalentDose......................................................................................53

Equation3:EffectiveDose.........................................................................................53

Equation4:IECExposureIndex.................................................................................55

Equation5:IECDeviationIndex................................................................................56



20

§



21

AbbreviationIndex

AAPM AmericanAssociationofPhysicistsInMedicine

AEC AutomaticExposureControl

ALARA AsLowAsReasonablyAchievable

AP Antero-Posterior

APDH AssociaçãoPortuguesaParaoDesenvolvimentoHospitalar

APIC AssociaçãoPortuguesadeIntervençãoEmCardiologia

ARS AdministraçãoRegionaldeSaúde

ASRT AmericanSocietyofRadiologicTechnologists

AUC AreaUndertheCurve

BaFX:Eu2+ BariumFluorohalideactivatedWithEuropium

BEIR BiologicalEffectsofIonizingRadiation

BMI BodyMassIndex

BSF BackscatterFactor

CaWO4 CalciumTungstate

CHLC CentroHospitalardeLisboaCentral

CHP CentroHospitalardoPorto

CHUC CentroHospitalareUniversitáriodeCoimbra

CIRSE CardiovascularandInterventionalRadiologicalSocietyofEurope

CPD ContinuousProfessionalDevelopment

CR ComputedRadiography

CsI CesiumIodide

CT ComputedTomography

DAP DoseAreaProduct

DDM2 DoseDataMed2

DI DeviationIndex

DICOM DigitalImagingandCommunicationsinMedicine

DQE DetectiveQuantumEfficiency



22

DR DigitalRadiography

DRLs DiagnosticReferenceLevels

E EffectiveDose

EC EuropeanCommission

ECSC EuropeanCoalandSteelCommunity

EEC EuropeanEconomicCommunity

EFOMP EuropeanFederationofOrganisationsforMedicalPhysics

EFRS EuropeanFederationofRadiographerSocieties

EI ExposureIndex

EIt TargetExposureIndex

EMDD EuropeanMedicalDeviceDirective

ESAK EntranceSurfaceAirKerma

ESPR EuropeanSocietyofPaediatricRadiology

ESR EuropeanSocietyofRadiology

EU EuropeanUnion

ExT ExposureTime

FPD Flat-PanelDetectors

FS-S FloorStandStandard

FSD Focus-SkinDistance

Gd2O2S:Tb Terbium-DopedGadoliniumOxysulfide

GDP GrowthDomesticProduct

Gy Gray

HR HumanResources

HVL HalfValueLayer

IAEA InternationalAtomicEnergyAgency

IAK IncidentAirKerma

ICRP InternationalCommissiononRadiologicalProtection

ICRU InternationalCommissiononRadiationUnits

ID Identification

IEC InternationalElectrotechnicalCommission



23

IP ImagePlate

IR InterventionalRadiology

KAP KermaAreaProduct

KSC Knowledge,SkillsandCompetences

kV TubeVoltage

LaOBr:Tm Thulium–DopedLanthanumOxybromide

LNT LinearNo-Threshold

mA TubeCurrent

mAs TubeCurrentTimeProduct

MED MedicalExposureDirective

MITA MedicalImagingandTechnologyAlliance

MRI MagneticResonanceImaging

MTF ModulationTransferFunction

NHS NationalHealthService

NPS NoisePowerSpectrum

OD OpticalDensity

OECD OrganisationforEconomicCo-OperationandDevelopment

P75 75thPercentile

PA Postero-Anterior

PACS PictureArchivingandCommunicationSystem

PHE PublicHealthEngland

PiDRL EuropeanDiagnosticReferenceLevelsforPaediatricImaging

PSL PhotostimulatedLuminescence

QA QualityAssurance

RHA RegionalHealthAuthority

ROC ReceiverOperatingCharacteristic

ROI RegionsofInterest

S/F Screen/Film

SD StandardDeviation

SI InternationalSystemofUnits



24

SID SourceImage-ReceptorDistance

SNK Student-Newman-Keuls

SNR Signal-To-NoiseRatio

SSD SourceSkinDistance

STUK FinnishRadiationandNuclearSafetyAuthority

Sv Sievert

TFT Thin-FilmTransistors

TLD ThermoluminescentDosimeters

UNSCEAR United Nations Scientific Committee on the Effects of AtomicRadiation

VGC VisualGradingCharacteristic

VS VerticalStand

WHO WorldHealthOrganization

WR Radiation-WeightingFactor

WT Tissue-WeightingFactor



25

IntroductionandobjectivesAccording to 97/43/EURATOM (Medical Exposure Directive - MED) Directive thepromotion and establishment ofDiagnostic Reference Levels (DRLs) ismandatoryforEUmemberstates.InPortugaltheDirectivewastransposedintonationallawbydecree-law 180/2002, 8 August. Evidence shows significant differences in dailyradiologicalpracticeatEuropean,nationalandhospitallevels,withobviousimpactonthecollectiveeffectivedosereceivedbythepopulation.

DatafromEuropeancountriesshowsawidevariationincommonDRLs,whichmaybe due to differences in socio-economic conditions, regulatory regimes, level ofactivity of professional bodies and in the structure of health care systems(private/public mix). International radiation protection bodies such as theInternational Atomic Energy Agency (IAEA) and the International Commission onRadiological Protection (ICRP) recommend that each country should carry out itsownnationalDRLsurvey(Edmonds,2009).

Researchers question whether there is any justification to explain the use of anexposure that is 10, 20 or even 126 times higher than that used by anotherinstitutiontoobtainsimilardiagnosticimages(Grayetal.,2005).Publishedstudies(Carroll&Brennan, 2003; Johnston&Brennan, 2000) reportedwide variations inpatient doses for the same radiographic examinations among hospitals in theUnitedKingdom.Thesevariationsareattributabletoawiderangeoffactorssuchastype of image receptor, exposure factors, number of images, type of anti-scattergridandlevelofqualitycontrol.

ThePortuguesehealthand/orradiationprotectionauthoritieshavenevertakenanykindofformalactiontodefineDRLs,neitherbyadoptingtheexistingonesfromtheEuropean guidance documents, nor by defining DRLs through surveys at nationallevel.

In fact the DRL concept, the need for optimisation and radiation protection inPortugal has only started to be knownand to be discussed in the last five years,through research activities driven by higher education institutions in radiographyandresearchcentresincombinationwithradiologydepartments.

There are published Portuguese National DRLs for paediatric head and chest CT(Santos,Foley,Paulo,McEntee,&Rainford,2014),howeverthereisaneedfortheofficialregulatoryauthoritiestoadoptandimplementthem.

The first known study developed in Portugal in the field of paediatric radiologyoptimisationresultedina70%reductiononEntranceSurfaceAirKerma(ESAK)andexposuretimeforpaediatricchestX-ray,afteratransitionfromscreen/film(S/F)toComputedRadiography(CR)systems(Paulo,Santos,Moreira,&Figueiredo,2011).



26

Thefindingsofthisstudywerethemotivationforthedevelopmentofthisthesis,astheyraisedseveralresearchquestions:

• Whattypeofpracticeisbeingusedforpaediatricplainradiography?• How do the exposure parameters influence patient dose exposure and

imagequalityinpaediatricimaging?• What is the impact of an optimisation programme in paediatric patients’

exposure?

Theseresearchquestionswillbeaddressedwithintheframeworkofthisthesis.

Toachievethis,amajorandseveralspecificobjectiveshavebeendefined:

Majorobjective:

ObtainDRLsforpaediatricplainradiography.

Specificobjectives:

• MeasureandevaluateKAPandESAK in themost frequentpaediatricplainradiographyproceduresandderivenumericvaluesofDRLs;

• Compare the obtained results with the “European guidelines on qualitycriteria for diagnostic radiographic images in paediatrics” and otherpublishedresults;

• Optimiseexamproceduresinordertoimproveradiographers’bestpractice;• Re-evaluate DRLs after optimisation actions and analyse the impact on

patientdose;• Develop amethodology to decrease radiation exposure in children, when

feasible.



27

Figure1showsthedesignstructureofthestudydevelopedinthisthesisinordertoaccomplishthedefinedobjectives.

Figure1:Schematicmapofresearchactivityandphasesoftheoverallthesis



28

§



29

1 BackgroundTechnological evolution and new scientific developments have driven the healthcare sector towards an unprecedented increase of its organisational complexity.One of the major contributors to that increase was, without doubt, thedevelopmentofmedicalimagingtechnology.

After Roentgen presented his manuscript, “On a new kind of Ray, A preliminaryCommunication”, to theWurzburgPhysicalMedicalSociety in1895, radiologyhastransformed itself froma scientific curiosity tooneof themainpillarsofmodernhealth care, becoming one of the scientific areas that contributed significantly totheunderstandinganddealingwiththedisease(Gagliardi,1996).Sincethatspecialmoment,theradiologybodyofknowledgehasbeenconstantlydeveloping,drivenbyapermanenttechnological(r)evolutionandisnowintegratedinalargespectrumofmedicalimagingprocedures(Lança&Silva,2013).

It is interesting to observe that 120 years after the revolution triggered byRoentgen’sdiscoveryofmedicalimaging,therearestillpersistingproblems,similarto those described in 1910 by Eddy German, one of the pioneers of theRadiographerprofessionintheUnitedStates:“Itwasdifficulttofindtwooperatorswho were anywhere near in accord regarding technical procedure. Some wouldadvisecertainproceduresandothersentirelydifferentprograms”(Terrass,1995).

Despitethescientificknowledgeandthetechnologicaldevelopmentinthepast120years,therealitydescribedbyEddyGermanin1910stillappliestotoday’spracticeof medical imaging. The reasons are manifold: (a) the lack of harmonisation ofprofessional practice at all levels; (b) a communication gap between science andprofessional practice; (c) a delay in integrating the new technology concepts ofmedical imaging into curricular programmes of health professions; (d) a barrierbetweenmanufactures/equipmentdevelopersandclinicalpractice.



30

§



31

1.1 PortugueseHealthcareContext

According to thePortugueseNational InstituteofStatistics (www.ine.pt),Portugalhas10,427,301 inhabitants,ofwhich19.8%are less than19yearsold (data from2013).

ThePortuguesepopulationhasaccesstoahealthcaresystemthatischaracterisedby three coexisting, overlapping systems: the national health service (NHS), auniversal, tax-financed system; public and private insurance schemes for certainprofessions (which are called health subsystems); and private voluntary healthinsurance.Thus,thePortuguesehealthcaresystemhasamixofpublicandprivatefunding. The NHS, which provides universal coverage, is predominantly fundedthrough general taxation. The health subsystems, which provide healthcarecoverage to between 20 and 25 per cent of the population, are funded mainlythrough employee and employer contributions (including contributions from thestateastheemployerofpublicservants).Closeto20%ofthepopulationiscoveredby voluntary private health insurance. About 30% of total expenditure onhealthcare is private, mainly in the form of out-of-pocket payments (both co-paymentsanddirectpaymentsbythepatient),andtoalesserextent,intheformofpremiumstoprivate insuranceschemesandmutual institutions(Barros&Simões,2007).

Portugalhasahealthexpenditureof10.2%ofitsGrowthDomesticProduct(GDP),abovetheaveragevalue(9.3%)oftheOrganisationforEconomicCo-operationandDevelopment(OECD)countries(OECD,2013).Howeverthisindicatorrepresentstheeffortthatthepopulationmakestohaveaccesstothehealthcaresystem.

Newmedicaltechnologies,suchasdigitalradiography(DR),computedtomography(CT)andmagneticresonanceimaging(MRI)areimprovingdiagnosisandtreatment,butarealsoincreasinghealthexpenditure(OECD,2013).

Considering the decrease of the Portuguese GDP in the last 5 years, the healthexpenditure per inhabitant has obviously decreased. Nevertheless Portugalpresentsbetterhealth indicators than themajorityof theOECDcountries. It isofinterestthatPortugalhasoneofthelowestinfantmortalityrates:3.4deaths/1000births(OECD,2015).

According to article 64 of the Portuguese Constitution (Assembleia da República,2005),theNHSispublicandprovidesuniversalcoverage.

The NHS, although centrally financed by the Ministry of Health, has a strongRegional Health Authority (RHA) structure since 1993, comprising five healthadministrations (AdministraçãoRegionaldeSaúde–ARS):ARSNorte,ARSCentro,ARSLisboaeValedoTejo,ARSAlentejoandtheARSAlgarve.



32

IneachRHAahealthadministrationboard,accountabletotheMinisterofHealth,manages theNHS. Themanagement responsibilities of theseboards are amix ofstrategicmanagement of population health, supervision and control of hospitals,and centralised direct management responsibilities for primary care/NHS healthcentres. The RHAs are responsible for the regional implementation of nationalhealthpolicyobjectivesandforthecoordinationofalllevelsofhealthcare(Barros&Simões, 2007). This organisation structure does not includeMadeira and Azores,since they have a special autonomous statute, however with the obligation tofollowandrespectthePortugueseConstitution.

Figure2:PortuguesemapindicatingtheRegionalHealthAuthorities(RHA).Numberofpatientsandhumanresources(HR)ineachRHA(ACSS,2015).

In Portugal a patient is classified as paediatric until 18 years of age (AltoComissariadodaSaúde,2009).Healthcareservicestothepaediatricpopulationcanbeprovided inanyhealthcarecentrethroughoutthecountry.However, therearethreededicatedpaediatrichospitals inPortugal:HospitalMariaPiadoPorto (ARSNorte); Hospital Pediátrico de Coimbra (ARS Centro); Hospital de D. Estefânia deLisboa (ARS Lisboa e Vale do Tejo). These dedicated paediatric hospitals haverecently been integrated into major hospital centres, respectively: CentroHospitalardoPorto(CHP),CentroHospitalareUniversitáriodeCoimbra(CHUC)andCentroHospitalardeLisboaCentral(CHLC).

These threemajor centres serve as reference hospitals for paediatric patients inPortugal, who need access to differentiated healthcare in all medical fields. Thethree centres have practitioners exclusively dedicated to paediatrics and are ingeneralequippedwithup-to-datetechnology.



33

1.2 EuropeanandPortugueselegalframeworksonionisingradiation

In1957,sixfoundingStates(Belgium,France,Germany,Italy,LuxembourgandtheNetherlands) joined together to form the European Atomic Energy Community(Euratom) and signed the Euratom Treaty in Rome. The main objective of theEuratom Treaty is to contribute to the formation and development of Europe'snuclearindustryandtoensuresecurityofsupply.

Before the European Community was founded, there had been the FoundingTreaties: European Coal and Steel Community (ECSC), European EconomicCommunity(EEC)andEuratom.In1967theywereallmergedtobecomelatertheEuropean Union.While the first two ended, Euratom is left unchanged and wasaddedasaprotocolonlytothenewEULisbonTreaty(EuropeanUnion,2007).

ThecurrentversionoftheEuratomTreaty(EuropeanCommission,2012)comprises177 articles, from which the articles quoted below are of relevance to medicalimaging:

• Article2:“…theCommunityshall…establishuniformstandardstoprotectthe health of workers and of the general public and ensure that they areapplied”;

• Article30:“BasicstandardsshallbelaiddownwithintheCommunityfortheprotectionof thehealthofworkersandthegeneralpublicagainstdangersarisingfromionisingradiations”;

• Article31:“ThebasicstandardsshallbeworkedoutbytheCommissionafterithasobtainedtheopinionofagroupofpersonsappointedbytheScientificand Technical Committee from among scientific experts, and in particularpublichealthexperts,intheMemberStates”.

Based on the Euratom Treaty the European Commission has published severalDirectives(bindinglegislationtobeimplementedbyEUMemberStates):Directives89/618/Euratom, 90/641/Euratom, 96/29/Euratom, 97/43/Euratom and2003/122/Euratom.

All these directives were repealed by the Council Directive 2013/59/Euratom(EuropeanCommission,2013a),withthemainobjectivestoconsolidatetheexistingEuropean radiation protection legislation into one document and to revise therequirementsoftheEuratomBasicSafetyStandards.

Accordingtoarticle106ofDirective2013/59/EURATOM,theMemberStatesshallbring into force the laws, regulations and administrative provisions necessary tocomplywiththeDirectiveby6February2018.ThereforeEuropeanMemberStateshave three years to adapt their national legislation to the new Europeanrequirements.



34

ThisthesisismainlyfocusedontheestablishmentofDRLs,aconceptintroducedbytheICRP(InternationalCommissiononRadiologicalProtection,1996)andadoptedfor the first time by the European Commission through the 97/43/EURATOMDirective (EuropeanCommission,1997)as:“dose levels inmedical radiodiagnosticorinterventionalradiologypractices,or,inthecaseofradio-pharmaceuticals,levelsof activity, for typical examinations for groups of standard-sized patients orstandardphantomsforbroadlydefinedtypesofequipment”.

With Directive 2013/59/EURATOM (European Commission, 2013a), the EuropeanCommissionmadeaclearprogressandstrengthenedtherequirementsinregardtoDRLs by changing the relevant text from: “Member States shall promote theestablishment and the use of diagnostic reference levels for radiodiagnosticexaminations”; to:“MemberStates shall ensure theestablishment, regular reviewand use of diagnostic reference levels for radiodiagnostic examinations”. Thereference toDRLs in the newDirective is related to optimisation and included inarticle56(EuropeanCommission,2013a).

AnotherimportantconceptincludedintheEURATOMDirectivesfromthebeginningis Clinical Audit, described as: “a systematic examination or review of medicalradiologicalprocedureswhichseekstoimprovethequalityandoutcomeofpatientcarethroughstructuredreview,wherebymedicalradiologicalpractices,proceduresand results are examined against agreed standards for goodmedical radiologicalprocedures,withmodificationofpractices,whereappropriate,andtheapplicationofnewstandardsifnecessary;”(EuropeanCommission,1997,2013a).

TheEuropeanCommissionrealisedthelackofunderstandingofhowtoimplementsuchanimportantconceptindailypracticeandpublishedtheGuidelinesonClinicalAudit forMedicalRadiologicalPractices(EuropeanCommission,2009)asatooltofacilitatetheachievementofitsgeneralobjectives:a)improvethequalityofpatientcare; b) promote the effective use of resources; c) enhance the provision andorganisationofclinicalservices;d)furtherprofessionaleducationandtraining.

Article18ofDirective2013/59/EURATOMisrelatedtoeducation,informationandtraininginthefieldofmedicalexposure.Accordingtothisarticle,“MemberStatesshallensurethatpractitionersandtheindividualsinvolvedinthepracticalaspectsofmedical radiological procedures have adequate education, information andtheoreticalandpractical trainingforthepurposeofmedical radiologicalpractices,aswellasrelevantcompetenceinradiationprotection”.

To give guidance to Member States the European Commission published theGuidelinesonradiationprotectioneducationandtrainingofmedicalprofessionalsin the European Union (European Commission, 2014a). These guidelines weredevelopedundertheMEDRAPETproject followingtheEuropeanKnowledge,Skillsand Competences (KSC) model (European Commission, 2008a) and represent an



35

important tool for education institutions, health authorities and regulatoryauthoritiestoimplementandauditeducationandtrainingprogrammesinradiationprotection.

Portugal is a member of the European Atomic Energy Community (Euratom)(EuropeanCommission,2012),theIAEAandtheOECDNuclearEnergyAgency.

As member of the Euratom Community, Portugal has to comply with the EUDirectives laying down basic safety standards for protection against the dangersarisingfromexposuretoionisingradiation.

ThereisnosingleframeworkactgoverningthenuclearsectorinPortugal.Instead,morethan100 laws,regulationsanddecreessetoutprovisionsgoverningnuclearactivities, frequently derogating each other implicitly, to the point where itbecomes amatter of doctrinal debate to identifywhich provisions are applicable(OECD,2011).

ThemainlegalinstrumentsgoverningtheuseofionisingradiationinhealthcareinPortugalare:

• Regulatory-DecreeNº.9/90,of19April,amendedbyRegulatory-DecreeNº.3/92,of6March,regulatingtherulesanddirectivesconcerningprotectionfromionisingradiation;

• Decree-Law Nº. 492/99, of 17 November, revised by Decree-Law No.240/2000,of26September,approvingthelegalframeworkforthelicensingand control of activities carried out in private health units using ionisingradiation, ultra-sound or magnetic fields for diagnostic, therapeutic orpreventive;

• Decree-Law Nº. 165/2002, of 17 July, amended by Decree-Law No.215/2008, of 10 November, setting out the competencies of the bodiesintervening in the field of protection against ionising radiation, as well asgeneralprinciplesofsuchprotection;

• Decree-Law Nº. 167/2002, of 18 July, amended by Decree-Law No.215/2008,of10November,settingoutthelegalframeworkforthelicensingandfunctioningofentitiesactiveinthefieldofradiologicalprotection;

• Decree-Law Nº. 180/2002, of 8 August, amended by Decree-Law No.215/2008, of 10 November, setting out the legal framework for theprotection of people’s health against the dangers arising from ionisingradiationinmedicalradiologicalexposures;

• Decree-Law Nº. 222/2008, of 17 November, setting out basic safety rulesconcerningthesanitaryprotectionofthepopulationandofworkersagainstdangersarisingfromionisingradiation.

SomeoftheselegalinstrumentswereimplementedaftertheEuropeanCommissiondecided to send a reasonedopinion to Portugal for failure to fulfil its obligations



36

relatedtobasicsafetystandardsfortheprotectionofthehealthofworkersandthegeneralpublicfromionisingradiation.

Portugal has notified transposition measures, which are dispersed in variouslegislative texts, instead of a coherent and consolidated legal framework. TheEuropeanCommissionconsidersPortugueselegislationonradiationprotectiontoocomplex, creatinguncertainty for thecitizens regarding the relevant transpositionprovisionsinforce(EuropeanCommission,2007a).

Atpresent,PortugalfulfilsalllegalrequirementsregardingthetranspositionoftheEuratomDirectives.However,althoughthemajorityandmostimportantaspectsofthebasicsafetystandardsforprotectionagainstthedangersarisingfromexposureto ionising radiation are legally published, the requirements are not observed indaily practice. It is not even possible to clearly identify, in the complex andentangled Portuguese legislation, who acts as the national radiation protectionauthority responsible forensuring the implementationof the radiationprotectionstandards.

Thetopicof this thesis isparticularly focusedonarticle61,nº1,paragraph (a)ofthe Directive 2013/59/EURATOM (European Commission, 2013a): “Specialpractices: Member States shall ensure that appropriate medical radiologicalequipment, practical techniques and ancillary equipment is used in medicalexposure:ofchildren”.



37

1.3 Theshiftofparadigminmedicalimaging

Theadvent,consequentdevelopmentandintegrationofcomputertechnologyintohealth industrywere the trigger for the conception of digital imaging systems inradiology, significantly improving imaging performance. Through an exponentialincreaseinthenumberofprocedures,medicalimagingdepartmentscontributedtoan unprecedented change in patientworkflow aswell as to enhanced diagnosticcapabilities. Throughout the world, and in particular in developed countries,conventional fluoroscopic and S/F radiography have been replaced by digitalsystems,CRorDR(Busch&Faulkner,2006).

Digitalradiographybroughtseveraladvantagestomedicalimagingprocedures,duetoitswidedynamicrange,thepossibilitytostoreandtransferimagesdigitallyandmost of all, due to the image post processing capabilities that the systems offer.However, despite all these technological features, with clear benefits for theworkflow of medical imaging departments, patient overexposure to ionisingradiationmightoccurwithoutvisibleimpactonimagequality.RadiographerswereusedtoS/Fsystemsthatwerebythemselvesa“self-control”system,sinceloworhigh exposures would deliver a “non diagnostic image” to the radiologistsresponsibleforimageinterpretation.Anoverexposedimagewastooblack,andanunderexposedimagewastoowhite.

The image processing algorithms of digital systems are standardised for eachimaging procedure. Therefore adequate gray-scale images are displayed correctlydespiteunderexposureoroverexposure(Donetal.,2013).Indigitalsystems,goodimages are obtained for a large range of doses (International Commission onRadiological Protection, 2004). Due to the fact that in digital systems there is aseparation between acquisition, processing and image display, a radiograph canhave an acceptable diagnostic quality, but could be under or overexposed(Herrmannetal.,2012).

Another reason related to the increase of patient dose, together with the widedynamic rangeof digital imaging systems, is the lack of training and/or the shortadaptationperiodtothenoveltechnologyinstalled.Veryoftendigitalsystemsareused in the samewayas S/F technology, including theuseof the sameexposurefactors,submittingpatientstohigherdoseswithoutbeingperceived(InternationalAtomicEnergyAgency,2011).Itmighthappenthatnon-optimisedexposurefactorsproducesuboptimum imageprocessing,hiding relevantdiagnostic information (C.Schaefer-Prokop,Neitzel,Venema,Uffmann,&Prokop,2008).

All these aspects combined with the lack of well-established methods to auditpatient doses in digital systems can increase the problem of patient radiationexposure(EliseoVañoetal.,2007).



38

Therefore a special attention must be given to continuous professionaldevelopment(CPD)ofradiographers,radiologistsandmedicalphysicists,inordertoensure that adequate knowledge, skills and competences are acquired whenmaking the transition from S/F radiography to digital systems (Nyathi, Chirwa, &VanDerMerwe,2010).

There is no doubt that the shift of paradigm inmedical imaging has significantlycontributedtoabetterandfasterhealthcaredelivery.However,thereareevidentchallenges for radiographers, radiologists, medical physicists, and other healthprofessionalsdirectlyinvolvedintheuseofionisingradiation,relatedtoadaptingtothisnewdigitalenvironment.Educationandtrainingaredefinitelyamongthemajorchallenges (European Commission, 2014a). This education and training process isevenmoreimportantinpaediatricradiology,wheremaintainingdiagnosticimagingqualitywithanachievablequalitydoseisevenmorecritical(Mooreetal.,2012).

Although recognising the potential of digital systems to improve the practice ofmedical imaging, the ICRP became aware of the risk of overuse of radiation. Tomanagethe identifiedrisks, the ICRPpublishedseveralspecificrecommendations,including appropriate training, particularly as regards patient dose management,revisionoftheDRLs,andfrequentpatientdoseaudits(InternationalCommissiononRadiologicalProtection,2004).

As faras thesituation inPortugal is concerned, there isnoofficialdata regardingthenumberandtypeofradiographyequipmentinstalled.Howeveraccordingtotheauthors’knowledgeandinformation,themajority(ifnotall)PortuguesepublicandprivatemedicalimagingdepartmentshaveshiftedtowardsCRand/orDRsystems.



39

1.4 PlainRadiographyDetectorSystems

AtraditionalX-rayimageisformedbydifferentshadesofgray(grayscale),eachonerepresentingthetissueandorgansX-rayattenuationproperties.ForagivenX-rayenergy,theattenuationcoefficientriseswiththeincreaseoftheatomicnumberoftheradiographedanatomicalstructure.

There are several events that occur when photons and electrons interact withmatter,suchasattenuation,absorptionandscattering,transformingtheenergyoftheoriginalprimaryX-raybeam.AnX-rayimageisthereforemadefromthecaptureof the energy output after the interaction of X-ray with the matter. The energycaptureprocesscanbedonebyusing(1)ascreen-filmor(2)adigitalsystem.

1.4.1 Screen-filmSystems

FordecadesS/Fsystemsincombinationwithvariousintensifyingscreenshavebeenthestandard formedical imagingbecauseof their functionalutilityandperceivedhigh-imagequality,andhavebeenusedtocapture,display,storeandcommunicatemedicalimaging.

AS/Fsystem(figure3)iscomposedof:

1. Cassette: a flat, light-tight container in which X-ray films are placed forexposure to ionising radiationandusuallybackedby lead toeliminate theeffectsofbackscatterradiation;

2. Intensifying screens: a plastic sheet coated with fluorescent material(phosphors),whichconvertsphotonenergytolight.

3. Film: consists of an emulsion-gelatin containing radiation sensitive silverhalide crystals, such as silver bromide or silver chloride, and a flexible,transparent,blue-tintedbase

Thetwomainphosphorsusedintheintensifyingscreensare:a)calciumtungstate(CaWO4),alsoknownasslowscreensduetotheirlowerefficiency,emittinglightinthe deep blue; b) rare earth phosphors, such as the terbium-doped gadoliniumoxysulfide (Gd2O2S:Tb), emitting green light, or thulium–doped lanthanumoxybromide (LaOBr:Tm)emittinggreen light (InternationalAtomicEnergyAgency,2014).RareearthphosphorsaremoreefficientatconvertingX-raystovisiblelightandthusfurtherreducetheradiationtothepatient.



40

Figure3:Screen-filmreceptor

(A)Openedcassetteshowingplacementoffilmandpositionofscreens,and(B)cross-sectionalviewthroughadualscreensystemusedingeneralpurposeradiographywiththefilmsandwiched

betweentwoscreens.

TheuseofintensifyingscreensdecreasestheabsorbeddosereceivedbythepatientcomparedtoX-raysdirectlyexposingthefilm.Filmsaretypicallyexposedby95%to99% light and to 1% to 5% of X-ray photons when intensifying screens are used(Bushong,2012).

TheemulsionofanexposedsheetofX-rayfilmcontainsthelatentimage.Althoughitlooksthesameasthatoftheunexposed,theexposedemulsionisalteredbytheexposure to light. The latent image is recorded as altered chemical bonds in theemulsion, which are not visible. The latent image is rendered visible during filmprocessingbychemical reductionof thesilverhalide intometallicsilvergrains,bychemicalprocessinginafilmprocessor(Lima,2009).

Althoughtheuseof thescreen-filmasadetectorsystemisbecomingobsoleteallaround the world and even not existing any more in most European countries(Portugal has no public or privatemedical imaging department using screen-filmsystems,althoughofficialinformationislacking),itisimportanttoanalysesomeofthe main features of this system, especially the denominated film characteristiccurve,alsoknownastheHurterandDriffield(H&D)curve,aplotofafilm’sopticaldensity(OD)asafunctionofthelogexposure.



41

Figure4:AHurtherandDriffieldcurve

TheregionsoftheH&Dcurveincludethetoe,thelinearregionandtheshoulder.Thebase+fogdensitycorrespondstotheODoftheunexposedfilm;adaptedfrom(Sprawls,2015)

Whena film isexposedtothe light froman intensifyingscreen, its response,asafunctionofX-rayexposure,isnonlinearandthecurvehasasigmoid(S)shape.Thetoe is the low-exposure region of the curve (meaning that less radiation andconsequently light reached that area of the film e.g.: bone, mediastinum, etc.).Between the toe and the shoulder of the curve is where ideally most of theradiographicimageshouldbeexposed.Beyondtheshoulderaretheareasofover-exposure (Lima,2009). It iseasy tounderstand thatwitha screen-filmmodel theresponseofthesystemislimitedtotheslopebetweenthetoeandtheshoulderofthe curve, and therefore has a limited dynamic range, which leaves theradiographerwithavery lowmarginoferrorwhenmakingradiographicexposurewithdiagnosticimagequality(Haus,1996).



42

§



43

1.4.2 DigitalSystems

Thedevelopmentofcomputer technology in thethirdquarterof theXXthcenturyledtoadramaticchangeintheorganisationalstructuresofoursociety,especiallyinthedevelopedcountries.Althoughthe impactwastransversal inallsectorsofthesociety, theadvancementof computer technologycreatedamajor (r)evolution inthehealthcaresector,particularlyinmedicalimaging.

Accordingto literaturetherearevariousdifferenttaxonomyapproachestodefineplainradiographydigitalsystems(Lança&Silva,2013),mainlyduetothefactthatseveral technologies were introduced in themarket in a very short time period,whichdidnotallowaconsolidationofconceptsanddefinitions.

Looking back in time and considering the technological features of plainradiography digital systems, the authors opted to use the taxonomy that splitsdigital systems inCRandDR (Korneretal.,2007;C.M.Schaefer-Prokop,DeBoo,Uffmann,&Prokop,2009).

Figure5:Taxonomyforplainradiographydigitalsystems

(Korneretal.,2007)

The introductionofCRsystems inmedical imagingdepartments in1983triggeredthe transition from screen-film to digital environments (Cowen, Davies, &Kengyelics,2007)andheraldedtheendofthetraditionalX-rayfilm.

Forthefirsttimearadiographcouldbedisplayedandviewedatseveralplacesbydifferentpersonsatthesametime,owingtothedevelopmentandimplementationofthePictureArchivingandCommunicationSystem(PACS)indailyroutine.

CRtechnologyisstillthemostwidelyuseddigitalacquisitionmethod,mainlyduetothe fact that it allows the transition fromS/F todigital systemswithout replacingtheinstalledradiographyequipment.



44

ACRsystemiscomposedofanimageplate(IP),aCRreaderandaviewingstation.TheIPismadeofathinlayerofphosphorcrystalsimplantedinabinderandfixedon a plastic substrate. The most frequently used phosphor material is bariumfluorohalide activated with europium (BaFX:Eu2+: where X represents one of thehalogensused,bromine(Br),iodine(I)orchlorine(Cl)atoms)(Cowenetal.,2007).Although appearing quite similar to a regular intensifying screen, an IP functionsquitedifferently.

Both intensifying screens and imaging plates rely on the principle of electronexcitation. Intensifying screens use a rare earth phosphor,which is a fluorescentmaterial thatemits lightphotonsafterbeing stimulatedbyX-rays.Thesephotonsareconvertedtoa latent imageonthefilmusingsilverhalidecrystalcentresasastoragemedium. The IP uses a phosphorescencematerial (BaFX:Eu2+) that,whenexposedtoX-rays,formsalatentimagedirectlyontheimagingplateitself,becausetheelectronsofthescreenareexcitedtoahigherenergylevelandaretrappedinhalide vacancies. Holes created by the missing valence electrons cause Eu2+ tobecomeEu3+.

Thistrappedenergycanbereleasedifstimulatedbyadditionallightenergyoftheproper wavelength by the process of photostimulated luminescence (PSL)(American Association of Physicists in Medicine, 2006). This latter process takesplaceintheCRreader.

Oncetheplateisinsidethereader,thephosphorisscannedwitharedlaserbeam,releasingthetrappedelectrons(atahighenergylevel),thatemitlightwhengoingback to their normal level of energy (Lança & Silva, 2013). The emitted light iscollected by a photodiode and converted into an electric signal to produce thedigitalimage(figure6).

Figure6:SchematicrepresentationofaCRreadersystem

AlaserbeamscanstheCRIPandreleasesthestoredenergyasvisiblelight.Aphotomultipliertubeconvertsthelighttoanelectricsignal.Aconvertercreatesthedigitalimage,whichisthensenttothe

computersystem.Adaptedfrom(InternationalAtomicEnergyAgency,2014)



45

Aftertheplateisscannedinsidethereader,itisexposedtoanintensewhitelighttoeraseit.ThisensuresthatanyresidualimageontheIPsiserased.IP’scanbereusedatleast10,000timesbeforetheyneedtobereplaced.

AlthoughtheimplementationofCRsystemshasallowedthechangeoverfromplainradiography to the new digital environment, the examination workflow has notchangedmuch.Radiographersstillhaveto:

• choosethesizeoftheIPaccordingtorequestedprocedure;• carry the IP to theX-rayequipmentandput it in the rightpositionon the

potter-bucky;• removetheIPafterexposureandtransportittotheCRreader.

DR systems were introduced in the market in the late 1980’s early 1990’s andimmediatelycreatedhighexpectations,sincetheintroductionofthenewflat-paneldetectors(FPD)promisedtosignificantlyimprovepatientworkflowbydramaticallydecreasingtheradiographicproceduretimeandtheradiographerworkload.IthasbeenshownthatbyintegratingtheFPDsystemsintodailypractice,productivityhasbeen further enhanced, since the IP manipulation step has been eliminated(Dackiewicz,Bergsneider,&Piraino,2000).

OneofthekeydifferencesbetweenFPDandCRisthefactthatFPDhaveadirectreadoutmatrixmadeofamorphoussilicon(aSi) thin-filmtransistors (TFT) (aSi-TFTelements).ThisTFTlayerisdirectlyattachedtoanX-rayabsorptionmedium(C.M.Schaefer-Prokopet al., 2009) and therefore thedigital image is directly sent to amonitordisplayimmediatelyaftertheexposure.

As shown in figure 7 there are two types of DR systems: a) those that use ascintillator (normally Cesium Iodide – CsI, doped with Thallium-Tl or GadoliniumOxysulfide-Gd2O2S)astheabsorptionmedium,whichtransformsX-rayintovisiblelight,thatisthencapturedbyaphotodiodeorbyaCoupleChargeDevice(CCD)oraComplementary Metal Oxide Semiconductor (CMOS): indirect conversion; or b)those that use a condensator material (normally amorphous selenium – aSe)attached directly to the TFT array: direct conversion, where the absorbed X-rayenergyisdirectlyconvertedintocharge,obviatingtheneedtohaveanintermediatesteptransformingX-rayintolight(Korneretal.,2007).



46

Figure7:SchematicrepresentationofDRsystems

Adaptedfrom(Chotas,Dobbins,&Ravin,1999)

All the systems described are available in the market, with the Food and DrugAdministration(FDA,UnitedStatesofAmerica)510(k)clearanceapprovalandwiththe CE mark, according to the European Medical Device Directive (EMDD)(European Commission, 2007b). It is important to note that both FDA (USGovernment, 2007) and EMDD do not require clinical trials for the pre marketauthorizationofmedicalimagingequipment.Therequirementstakenintoaccountaremainlyrelatedtoqualityandsafetyspecifications.

Taking that into consideration, vendors are free to offer any type of DR systemapproved for the market. Therefore it is important to understand the differentcharacteristics of each system that, depending on thematerial used,will lead todifferentphysicalperformanceofthedetector.

DetectiveQuantumEfficiency(DQE)iscurrentlyestablishedasthegoldstandardtomeasure the detector performance (Lança & Silva, 2013). When assessing thephysical efficiency of a radiological digital detector, the measurement of imagequality (using Signal-to-NoiseRatio - SNR)mustbe referred to the radiationdoseusedtocreatetheimage.Ingeneralterms,effectiveradiographicimagingdemandsthemaximisationofrecordedSNR,whileminimisingtheradiationdosedeliveredtothedetector(Cowenetal.,2007).

AgoodimagingdetectorintermsofitsnoiseperformanceisonethatproducesanoutputsignalwiththesameSNRasitsincomingsignal,i.e.doesnotaltertheSNR.Itisdifficult,ifnotimpossible,toimproveSNRwithoutdegradingsomeotheraspectofsystemperformance(C.M.Schaefer-Prokopetal.,2009).

By definition, if a detector receives data with an SNR of SNRin, from which itproducesdatawithaSNRofSNRout,thentheDQEofthedetectoris:

Equation1:DetectiveQuantumEfficiency



47

Anideal(butunrealistic)detectorpreservestheSNR,recordingeveryincidentX-rayquantum,andwhichthereforehasaDQEof1.DQEdetectoralways lies in therange0>DQEdetector<1.However,alldetectorsavailableonthemarkethaveaDQEalwayslowerthan1(Chotasetal.,1999).

No imagedetectorcanabsorball the incidentX-rayphotonswith100%efficiency(Cowen et al., 2007). Inevitably some X-ray photons pass through the detector,whilesomethatareabsorbedmaybere-emittedandexitthedetector.Thislossininformation carriers is compounded by secondary losses due to the presence ofextraneousnoisesourcesintheimagedetectoritself.

Forradiographyapplicationswithahighdosedeliveredtothedetector,bothdirectandindirectconversionflat-paneldetectorshaveahigherDQEthanS/FsystemsorCRsystems(Kotter&Langer,2002).

The International Electrotechnical Commission (IEC) publisheda standardmethodformeasurementoftheDQE(InternationalElectrotechnicalCommission,2015)thatalsoincludedspecificationsforthemeasurementoftwoassociatedmetricsthatareimportanttocharacteriseadetector:theModulationTransferFunction(MTF)andtheNoisePowerSpectrum(NPS)(McMullan,Chen,Henderson,&Faruqi,2009).

TheMTF is a measure of the ability of an imaging detector to reproduce imagecontrast from subject contrast at various spatial frequencies. At a given spatialfrequency, the value of theMTFwill lie between zero and one. AnMTF of zeromeansthatnosignalmodulationisbeingreproduced,andanMTFofoneindicatesa perfect transfer of the signal. Typically, a system's ability to represent a signaldecreasesasthespatialfrequencyofthesignalincreases(James,Davies,Cowen,&O’Connor,2001).

TheNPS, also knownasWiener Spectra (afterNorbertWienerwhopioneered itsuse),describesthenoisepropertiesofan imagingsystem(Park,Cho,Jung,Lee,&Kim,2009)andisametricofimagequality,providingamoredetaileddescriptionoftheoverallnoiseinanimage(Lança&Silva,2013).

KnowledgeoftheDQE,NPS,andMTFforDR,allowsobjectivecomparisonsthatcanassist in the determination of appropriate and reasonable performance for aparticular imaging application (American Association of Physicists in Medicine,2006).

DR techniqueshave thepotential to improve imagequality and, given thehighersensitivityoftheir imagereceptorscomparedwithS/F,alsoofferthepotentialfordose reduction. However, in practice, since image receptors also have a broaderdynamicrangethanfilm,higherdosesmayalsooccur(InternationalAtomicEnergyAgency,2007).



48

AsshowninFigure4(pag.41),thedynamicrangegradationcurveofS/Fsystemsis“S” shaped, within a narrow exposure range for optimal film blackening.Radiographerswereaware thatS/Fsystemshada lowtolerance foranexposure,resultinginfailedexposuresorinsufficientimagequality.

Withtheshifttothedigitaltechnologyenvironment,radiographersstartedtousedetectorswithawiderandlineardynamicrange,whichinclinicalpracticevirtuallyeliminates the risk of a failed exposure. Although the positive aspects of digitalsystemsclearlyprevail,radiographersneedtounderstand,thatspecialcarehastobetakennottooverexposethepatientbyapplyingmoreradiationthanisneededforobtainingadiagnosticqualityimage,becausethedetectorfunctionimprovesasradiationexposureincreases.

This phenomenon, described in literature as “exposure creep”, is defined as anunintendedgradualincreaseinX-rayexposuresovertimethatresultsinincreasedradiation dose to the patient when shifting from S/F to DR systems (Gibson &Davidson,2012).InDR,imageprocessingcancompensatebyupto100%forunderexposureandupto500%foroverexposure,andstillproduceaclinicallyacceptableimage(Butler,Rainford,Last,&Brennan,2010).

The described phenomenon can be easily seen in figure 8, (from Lança & Silva,2013)inwhich,duetothelinearsignalresponseofdigitalsystems,itispossibletohave comparable quality images with an exponential difference in patient doseexposure.

Figure8:DynamicrangeindigitalandS/Fsystems(Lança,2013)



49

To better understand this phenomenon and to maximise the potential of thetechnological featuresofthe imagingequipment,there isaneedforapermanentmultidisciplinaryapproach,involvingmedicalphysicists,radiographersandmedicaldoctors, todevelopharmonisedstandardsofpractice towards thebestdiagnosticimagequalityatthelowestdosepossible.

DR systems have an enormous potential for high image quality at lower doses.However, overexposed images can easily go unnoticed, resulting in unnecessaryoverexposureandpotentialharmtothepatient(Seibert&Morin,2011).Thereforeoptimisation should be used as a permanent process to reduce patient dose(Neofotistou,Tsapaki,Kottou,Schreiner-Karoussou,&Vano,2005).



50

§



51

1.5 Dosedescriptorsinradiography

Patient dosimetry in diagnostic imaging is complex and involves severaluncertaintiesduetothelargevariationoftechnologyandtechniques.

Theobjectiveofdosimetryinradiologicalimagingisthequantificationofradiationexposurewithinanapproach tooptimise the imagequality to theabsorbeddoseratio. The image quality should be understood as the relevant informationappropriatetotheclinicalsituation(InternationalCommissiononRadiationUnits&Measurements,2005).

Thereareseveraltermstodescriberadiationdoseandconfusioncanarisethroughtheinappropriateuseofitfromclinicalprocedurestopatients(Nickoloff,Lu,Dutta,&So,2008).

The key dosimetric quantities for use in general radiography and recommended(InternationalAtomicEnergyAgency,2013)forpaediatricpatientdoseare:

• incidentairkermaKa.i(IAK);• entrancesurfaceairkerma,Ka.e(ESAK);• airkerma(ordose)areaproduct,Pka(KAPorDAP).

These are the recommended application-specific dosimetric quantities for theimplementation of DRLs (International Commission on Radiation Units &Measurements,2005),whichhavethusbeenusedforthisthesis.

IAK is measured for phantoms and is determined using recorded exposureparametersforpatients.Forpatients,ESAKistypicallydeterminedfromtheIAKbyapplying the appropriate backscatter factor (BSF), but may also be measureddirectly with thermoluminescent dosimeters (TLD) or derived from the PkameasuredusingaKAPmeter(InternationalAtomicEnergyAgency,2013).TheBSFforradiographyrangefrom1.25to1.60fortypicalX-rayspectraandfieldsizesusedforadults(Petoussi-Henss,Zankl,Drexler,Panzer,&Regulla,1998).Thebackscatterfactordependson theX-rayspectrum, theX-ray field size,SIDand thepatientorphantomthickness.

According to International Commission on Radiation Units (ICRU) report 74(InternationalCommissiononRadiationUnits&Measurements,2005):

• TheIAK(Ka.i) isthekermatoairfromanincidentX-raybeammeasuredonthe central beam axis at the position of the patient or phantom surface(Figure 9).Only the radiation incident on the patient or phantomandnotthebackscatteredradiationisincluded.ItisexpressedinJ/kgandtheunitisGray(Gy);

• TheESAK(Ka.e)isthekermatoairmeasuredonthecentralbeamaxisatthe



52

positionofthepatientorphantomsurface(Figure8).Theradiationincidentonthepatientorphantomandthebackscatteredradiationareincluded.ItisexpressedinJ/kgandtheunitisGray(Gy).

• TheKAPorKAP(PKA)istheintegraloftheairkermaovertheareaoftheX-raybeaminaplaneperpendiculartothebeamaxis(Figure9).ItisexpressedinJ.kg-1.m2andtheunitisGy.m2.

Figure9:Schematicrepresentationofaradiographwithsomedosimetricandgeometricquantitiesfor

determinationofpatientdose.(ICRUreport74)

There are also ionising radiation risk-related dose quantities (InternationalCommission on Radiation Units & Measurements, 2005) that have beenrecommendedbyICRP,suchas:

• Exposure:ameasureofradiationbasedonitsabilitytoproduceionisationinairunder standard temperatureandpressure.The International SystemofUnits(SI)unitforexposureisCoulombs/kginair.

• Absorbed dose: the amount of energy absorbed per mass is known asradiationdose(DT).Radiationdoseistheenergy(Joules)absorbedperunitmassoftissueandhasthe(SI)unitsofGray(1Gy=1J/kg).

• Equivalent dose: the term ‘equivalent dose’ is used to compare thebiological effectiveness of different types of radiation to tissues. The (SI)doseequivalent(HT)inSievert(Sv)istheproductoftheabsorbeddose(DT)inthetissuemultipliedbyaradiation-weightingfactor(WR),oftencalledthequality factor (InternationalCommissiononRadiological Protection, 2012).Equivalent dose is expressed as a summation to include the effects of



53

irradiationoftissuebymorethanonetypeofradiation.

Equation2:EquivalentDose

• Effectivedose:EffectiveDose(E)isusedtoestimatetheriskofradiationinhumans. It is the sum of the products of equivalent doses to eachorgan/tissue (HT) and the tissue-weighting factor (WT) (InternationalCommissiononRadiologicalProtection,2012).Theunitofeffectivedose istheSievert(Sv).

Equation3:EffectiveDose

• Collectivedose:CollectivedoseisdefinedasthedosereceivedperpersoninSvmultipliedby thenumberof persons exposedper year i.e.man-Sievertper year. This unit is generally used for protection purposes and inpopulationresponsecalculations.

As one can see there are several dose quantities representing different conceptsbut yet using the same units. This can represent a problem amongst healthprofessionals, especially between radiographers, radiologists and medicalphysicists,creatingconfusionintheirdailypractice.Thereforeitcouldbebeneficialto develop a new system where each unit immediately brings to mind thecorrespondingquantitytowhichitrefers(Rehani,2015).

Although theexposure index (EI) it isnot consideredby ICRUand ICRPasadosedescriptor, it could be understood as an exposure concept related to the patientduetoitsinfluenceonimagequalityandthefactthatitisrelatedtothedelivereddoserequiredforanyradiologicalimage(Lança&Silva,2013).

For some time DR systems have associated the EI to the concept of S/F “speedclass”.Thishascreatedsomeconceptualmisunderstandingandscientificconfusionamongstusers(Huda,2005).

ItisimportanttonotethatpatientradiationdoseandEIarenotthesame.Thedoseto the patient is determined by the radiograph exposure technique factors (kV,mAs,grid,sourcetoimage-receptordistance(SID),filtration,beamcollimation),theX-raybeampenetrability andquality, aswell asby the size andareaof thebodyirradiated.TheEIonthedetectorisdeterminedbytheremnantradiation(primary



54

radiationtransmittedthroughthepatientandscatteredradiationtransmittedfromthepatient)thatisabsorbed,convertedtoelectronicsignals,andtransformedintoadigitalradiographicimage(Seibert&Morin,2011).

Currently each manufacturer has its own definition and reference values for EI(table 1). These EIs are entirely manufacturer-specific and vary greatly interminology,mathematicalformulasandcalibrationconditions.Moreover,insomesystems, increasing EI values indicate increasing dose whereas for others theoppositeisthecase.This inconsistencybetweenvendorshasbeenarestrictiontoaneffectiveuseofEIandconfusionispatentamongstprofessionalsthatusemorethanonesystemindailypractice(Mothiram,Brennan,Lewis,Moran,&Robinson,2014).

Table1:Manufacturerexposureindexnameandindicatorofdigitalsystems

Manufacturer Name Symbol ExposureIndicator

Agfa Logofmedianofhistogram

lgM

Logarithmic value (lGm): themedian value of theROIhistogramdefinesthelGmfortheimage. lGmis a deviation index as it compares the resultantlGmvaluetoareferencelGmvalue.

Canon ReachedExposureValue

REX REXisafunctionofthebrightnessandcontrastasselectedbytheoperator

Philips ExposureIndex EIEI is inversely proportional to the air kermaincidentontheimagereceptor.Itisderivedfromacharacteristicpixelvalueoftheimage.

Carestream ExposureIndex EI

EIisanumericalvaluecomputedfromtheaveragecode value of those areas of the image data thatare used by the image-processing algorithm tocompute the original tone scale. It has alogarithmicrelationshipwiththeairkermaincidentonthedetector.

Fujifilm Svalue SSisrelatedtotheamountofamplificationrequiredby the photomultiplier tube to adjust the digitalimage.Sisinverselyproportionaltoexposure.

GeneralElectric

DetectorExposureIndex

DEI DEI compares the detector exposure to theexpectedexposurevalue

Siemens ExposureIndex EXI

Theexposedfieldisdividedinto3X3Matrix,wherethecentralsegmentistheROI.EXIiscalculatedasthe average out the original pixel values in thecentral segment. EXI value is directlyproportionalto dose. Doubling of EXI value represents adoublingofabsorbeddoseatimagereceptor.



55

TheIECandtheAmericanAssociationofPhysicistsinMedicine(AAPM)havebeenworking separately on exposure value standardisation. Both efforts involvecollaboration among physicists, manufacturers, and the Medical Imaging andTechnology Alliance (MITA). IEC standard 62494-1 (International ElectrotechnicalCommission,2008)waspublishedin2008,andthereportofAAPMTaskGroup116(AAPM Task Group 116, 2009) was published in 2009.Even though theimplementationofthesestandardsisnotalegalrequirement,somemanufacturershave already adopted them and others will likely follow (Don, Whiting, Rutz, &Apgar,2012).

IndigitalX-rayimaging,anEIisusedtodescriberadiationdosetothedetector.TheEIisavaluabletoolintheoptimisationprocess,sinceindigitalimagingitispossiblethatthepatientgetsaninappropriatelyhighradiationdosewhiletheimagequalityisstillacceptable.

AccordingtotheIECstandard62494-1(InternationalElectrotechnicalCommission,2008) there are three new concepts that will be introduced in radiologydepartments:

1. EXPOSURE INDEX (EI): the EI is the measure of the detector response toradiationintherelevantimageregion(typicallytheregionforwhichtheexposureparametersshouldbeoptimised)andisdefinedas:

Equation4:IECExposureIndex

whereKCAL is thedetectorairkerma inµGyunderbeamqualityRQA-5calibrationconditions (International Electrotechnical Commission, 2005b). The physicalcharacteristicsofaRQA-5beamqualityare:

• addedfiltrationof23.5mmAl;• tubevoltageof70kV;• HalfValueLayer(HVL)of7.1mmAl

EI is not a patient dose indicator representing only a relative exposuremeasureaccordingtothetypeoftheexamination(Donetal.,2012).

2. TARGETEXPOSURE INDEX (EIT): the EIt is an expected value of the EIwhenexposing the image receptor properly and obtaining an optimalradiographicimage.

3. DEVIATIONINDEX(DI):theDIquantifieshowmuchtheEIvariesfromtheEItandisdefinedas:



56

Equation5:IECDeviationIndex

Ideally, EI andEIT shouldbe the sameand thereforeDIwouldbe0.However, anacceptableDIvalueisexpectedtobebetweena-0.5to0.5range.DIvaluesof1.0to3.0correspondto26%and100%ofoverexposure,respectively,whileDIvaluesof-1.0to-3.0correspondto20%and50%ofunderexposure,respectively.OneofthemajoradvantagesoftheDIvalueisthatitgivesanimmediatefeedbacktotheradiographerabouttheappropriatenessoftheexposure(Donetal.,2012).

The IEC standard is a helpful tool to eliminate the various different terms andconceptsimplementedbythevendorsandconsequentlyendtheexistingconfusionamongstradiographers,radiologistsandmedicalphysicists.

Since this is a new EI concept, there is a need to define proper EIt for themostcommonexaminations,takingintoconsiderationthediagnosticimagequality.

It is expected that in the near future all vendors implement this new IEC indexstandard in their equipment. It will be a major challenge for radiologists,radiographers and medical physicists (as individuals or through their scientificbodiesrepresentatives)todeveloptogetherwith internationalorganizations(suchasICRPandIAEA)EItreferencevaluesforthemostcommonprocedures,takingintoaccountthepatientcharacteristicsandtheclinicalindicationofthereferralfortheradiographicexposure.



57

1.6 Risksinpaediatricimaging

The risk of exposure to radiation is a permanent topic on the agenda of globalorganisationsliketheICRP(www.icrp.org),theUnitedNationsScientificCommitteeon the Effects of Atomic Radiation (UNSCEAR, www.unscear.org), the IAEA(www.iaea.org)andtheWorldHealthOrganization(WHO,www.who.int).Theroleof theseglobalorganisations is crucial, as theycontinuouslyevaluateandanalysethe scientific literature about the effects of ionising radiation and also publishrecommendationsandguidelinesonhowtouseionisingradiationinasafeway.

The 1990 recommendations of the ICRP defined two different types of potentialeffects on tissues and organs exposed to ionising radiation. Theserecommendations were formally replaced in 2007 (International Commission onRadiologicalProtection,2007)bythefollowing:

1. Deterministicortissuereactioneffectsarecharacterisedbyathresholddosebelow which they do not occur. Deterministic effects have a clearrelationshipbetweentheexposureandtheeffectandtheleveloftheeffectis directly proportional to the amount of the dose received. These effectsare often evident within hours or days. Examples of deterministic effectsincludeerythema(skinreddening),skinandtissueburns,cataractformation,sterility,radiationsicknessanddeath.

2. Stochastic effects are those that occur by chance and consist primarily ofcancer and genetic effects. Stochastic effects often occur years afterexposure. As the dose to an individual increases, the probability ofdevelopingcancerorageneticeffectalso increases.Forstochasticeffects,thereisnothresholddosebelowwhichitisrelativelycertainthatanadverseeffectwillnotoccur.

Stochasticrisksareofspecialconcerninpaediatricimaging,sincechildrenaremorevulnerablethanadultsandhavealongerlife-spantodeveloplong-termradiation-inducedhealtheffectslikecancer.

Itiswellknownthatthenaturalbackgroundisbyfarthehighestsourceofionisingradiation exposure for both children and adults and that there is significantgeographical variation in the doses received. However, there are no differencesbetween the doses received by children and adults in the same location. Theaverage annual effective dose to an individual resulting fromnatural backgroundradiationisapproximately2.4mSv(UNSCEAR,2013).

There is an on-going controversy among the scientific community about theacceptance of the Linear No-Threshold (LNT) model. The LNT model hypothesisassumes that the long term, biological damage caused by ionising radiation(essentially the cancer risk) is directly proportional to the dose, and that any



58

increment of exposure above natural background levels will produce a linearincrementofrisk.TheLNTmodelhypothesishasothercompetingtheories:a)theThreshold Model: very small exposures are harmless; b) the Radiation HormesisModel:radiationatverysmalldosescanbebeneficial(Walletal.,2006).

TheFrenchAcademiesconsiderthattheLNTmodelforassessingcarcinogenicrisksinducedbylowdoses,suchasthosedeliveredbydiagnosticradiology,isnotbasedon valid scientific data andmight create anxiety amongst patients, although theyacknowledge that the model can be practical for the organisation of radiationprotection(Tubiana,2005).

A different opinion is supported by the ICRP and by the National Academy ofSciences(BiologicalEffectsoflonizingRadiation-BEIRVIIreport)thatrecommendstheuseoftheLNTmodel(Tubiana,Feinendegen,Yang,&Kaminski,2009).

TheBEIRVIIcommitteeconcludesthatcurrentscientificevidenceisconsistentwiththehypothesisthatthereisalineardose-responserelationshipbetweenexposureto ionising radiation and the development of radiation-induced solid cancers inhumans(CommitteetoAssessHealthRisksfromExposuretoLowLevelsofIonizingRadiation,2006).

As statedabove, radiation isoneof themostextensively researchedcarcinogens,but the effects of low doses are still somewhat unclear. The weight of evidencefrom experimental and epidemiological data does not suggest a threshold dosebelow which radiation exposure does not cause cancer. If there is no suchthreshold, then diagnostic X-rays are likely to induce some cancers (González &Darby,2004).

An increased risk of acute lymphocytic leukaemia from plain radiography and offatal breast cancer from scoliosis series in children has been demonstrated andreported by some authors (Willis& Slovis, 2005). The younger the patient at thetimeofexposure,thegreatertheriskofdevelopingafatalcancer.

Despite all the controversies found in the literature, the LNT model hypothesisremains a wise basis for radiation protection at low doses and low dose rates(International Commission on Radiological Protection, 2005) and represents, untiltoday,thebestmeansforradiation-protectionstandards(Breckow,2006).

Thebiologicaleffectsofionisingradiationinchildrenarebasedonaveryimportanttrilogy:1)radiosensitivity;2)lifeexpectancy;3)radiationexposure.Consideringthisit is expected that for the sameeffectivedose, thebiological effects and lifetimerisks are higher in children than in adults (International Atomic Energy Agency,2012).

An UNSCEAR report from 2013 reviewed 23 different cancer types. Broadly, forabout 25 per cent of these cancer types, including leukaemia and thyroid, skin,



59

breastandbraincancer,childrenwerefoundtobeclearlymoreradiosensitive.Forsomeofthesecancers,dependingonthecircumstances,theriskswerefoundtobeconsiderably higher for children than for adults. In diagnostic medical exposure,children may receive significantly higher doses than adults for the sameexamination,ifthetechnicalparametersfordeliveringthedosearenotspecificallyadapted.Cancerspotentiallyinducedbyexposuretoionisingradiationatyoungagemayoccurwithina fewyears,butalsodecades later.Estimatesof lifetimecancerrisk for persons exposed as childrenwere described as uncertain andmight be afactoroftwotothreetimeshigherthanestimatesforapopulationexposedatallages(UNSCEAR,2013).

The sizeandweightofpaediatricpatientshasabig impacton the radiationdosereceived.SmallerandlighterpatientshavelowerattenuationoftheprimaryX-raybeamandarethereforeexposedtoahigherradiationdose.Insmallerandthinnerpaediatric patients the organs are closer and therefore more easily exposed toscatteredradiation(Linet,Kim,&Rajaraman,2009).

A basic knowledge of radiation risk is useful in counselling patients who expressconcernaboutthis issue. Inmostcases, thebenefitsof indicatedmedical imagingwill outweigh the relatively small excess of cancer risk, and patientmanagementshouldnotbealteredonthebasisofradiationrisk.However,forcertainsubsetsofpatients,radiationriskshouldbeofgreaterconcerntotheclinician(Lin,2010).

Evenknowing that thereare several controversies regarding thequantificationofrisksandalsoalackofagreementonhowtopresentit,thebestattitudeforhealthprofessionals using ionising radiation in clinical practice is to perform a medicalimagingprocedureaccordingtotheALARAprinciple,accordingtowhichradiationdosesshouldbekeptAsLowAsReasonablyAchievable, taking intoaccountsocialandeconomic factors (InternationalCommissiononRadiologicalProtection,1977,2007).

Sometimestheabbreviation‘ALARA’isusedasequivalenttotheterm‘optimisationofprotection’or in replacement thereof.However, it shouldbekept inmindthatthe expression ‘as low as reasonably achievable’ is only part of the concept ofoptimisation.Theentireconceptimplies,moreprecisely,keepingpatientexposuretotheminimumnecessarytoachievetherequiredmedicalobjective(diagnosticortherapeutic). In diagnostic imaging and x-ray-guided interventions, it means thatthenumberandqualityof imagesaresufficienttoobtainthe informationneededfordiagnosisorintervention(InternationalCommissiononRadiologicalProtection,2013).



60

§



61

1.7 TheinternationalcontextofDiagnosticReferenceLevels

DRLs were introduced by ICRP as a practical tool for optimisation in diagnosticradiology and nuclear medicine (International Commission on RadiologicalProtection, 1996). Achieving acceptable image quality or adequate diagnosticinformation, consistent with the medical imaging task, is the overriding clinicalobjective.DRLsareusedtohelpmanagetheradiationdosedeliveredtopatientssothatthedoseiscommensuratewiththeclinicalpurpose.

DRLs should be used as a form of investigation level to identify unusually highlevels.IfDRLsareconsistentlyexceeded,alocalreviewusuallytakesplace.DRLsarenotintendedforregulatoryorcommercialpurposes,nordotheyrepresentadoseconstraint,noraretheylinkedtolimitsorconstraints(InternationalCommissiononRadiologicalProtection,2001).

The radiationprotectionschemeusedacrossEurope (andworldwide) isbasedonthe recommendations of the ICRP, which are founded on the dose descriptorsdefinedbytheICRU.ThepublicationICRP60(ICRP1991)recommendsaradiationprotectionsystembasedonthesystemofdoselimitation,whichhasalsobeenanessentialelement inearlier ICRPdocumentssuchas ICRP26 (ICRP1977). ICRP60(ICRP,1991)wassubstantiallyrevisedandupdatedin2007withthepublicationofICRP103(ICRP2007).

The ICRPsystemofradiationprotection isbasedonthreefundamentalprinciples:justification,optimisationanddoselimitation.

The principle of justification requires that any decision that alters the radiationexposure situation should do more good than harm; in other words, theintroduction of a radiation source should result in sufficient individual or societalbenefittooffsetthedetrimentitcauses.

Theprincipleofoptimisationrequiresthatthelikelihoodofincurringexposures,thenumberofpeopleexposedandthemagnitudeoftheir individualexposureshouldall be kept as low as reasonably achievable, taking into account economic andsocietal factors. In addition, as part of the optimisation procedure, the ICRPrecommends that there should be restriction on the doses to individuals from aparticularsource,whichhasledtotheconceptofdoseconstraints.

TheprincipleofdoselimitationrequiresthatthedosetoindividualsfromplannedexposuresituationsotherthanmedicalexposureofpatientsshouldnotexceedtheappropriatelimitsrecommendedbytheICRP.

AlsotheEuropeanCommission(EC)hasstrengthenedtheimportanceofDRLsintherecently published Council Directive 2013/59/EURATOM (European Commission,2013a), laying down basic safety standards for protection against the dangers



62

arisingfromexposuretoionisingradiation.TheECrecognisesthatthetechnologicaland scientificdevelopments in themedical fieldhave led toanotable increase inthe exposure of patients. Therefore the Directive should strengthen therequirementsconcerninginformationtobeprovidedtopatients,therecordingandreportingofdosesfrommedicalprocedures,theuseofDRLsandtheavailabilityofdose-indicatingdevices.

Article 56 of the Directive 2013/59/EURATOM defines that Member States shallensure the establishment, regular review and use of DRLs for radiodiagnosticexaminations,havingregardtotherecommendedEuropeanDRLswhereavailable.

The use of DRLs generates complementary information and thus supportsprofessional judgment. The use of DRLs is important to promote the review ofpractice in local sites. The establishment of DRLS at local level should beimplemented with the involvement of radiologists, radiographers and medicalphysicistsandshouldberegularlyreviewedtoimprovebestpracticeatlowerdoses(Santos,2014).

SinceICRPintroducedtheDRLconcepttothescientificandprofessionalcommunityin 1996, several initiatives were taken to try and make its use effective in dailypractice.OneofthemostrelevantstepswastheinclusionoftheDRLconceptintheEuratom97/43Directive,whichservedasatriggerforseveralotheractions,mainlynationalradiationprotectionauthorities.

Aware of the importance of using the DRL concept as a tool to decrease doseexposure to the patients and the population, the EC published several guidelinedocuments (European Commission, 1996a, 1996b) for both adult and paediatricplain radiography, with the main objective to achieve adequate image quality,comparablethroughoutEuropeatreasonablylowradiationdoseperradiograph.

AnotherimportantguidanceonDRLsformedicalexposurewaspublishedbytheECin1999 (EuropeanCommission,1999),highlighting the importanceofestablishingDRLs for high-dose medical examinations, in particular CT and interventionalradiology (IR) procedures and for patient groups that are more sensitive toradiation,especiallychildren.

The RP 109 document recommends that DRLs for diagnostic radiology should bebasedondosesmeasuredinvarioustypesofhospitals,clinicsandpracticesandnotonlyinwell-equippedhospitals.Thevaluesshouldrepresentthe75thpercentileofthe ESAK (mGy) and/or the KAP (Gy.cm2). According to the RP 109 guidancedocument,theKAPisamorepracticaldosedescriptor,sincetheentireexaminationisrecordedwithoutinterferingwiththepatientandtheexamprocedure.



63

Althoughextremelyimportant,thisguidancedocument(RP109)onlypresentsDRLsfor plain S/F systems, and is only of limited value regarding paediatric patients,since the DRLs are based on a standard sized five-year old patient. In fact thislimitation ismajor, since the ‘paediatric population’ includes the age range frombirthuntil18years(EuropeanCommission,2006),whichimpliesahugevariationinpatient weight, ranging from some grams to more than 100 kg. This enormousheterogeneity of patient characteristics, combined with the fact that pre-setsinstalled in radiological equipment are normally not adapted to them, makesdefiningDRLsandoptimisationofproceduresverychallengingforthispopulation.

Thereisalsoalackofconsensusintheliteratureregardingthebestmethodologytogroup paediatric patients in order to define a DRL. Some authors group patientsaccordingto:a)specificages(0,5,10and15years);b)divisionbetweennew-bornsand infants; c)agegroups (<1,1-<5,5-<10,10-<16,≥16).Otherauthors (Kiljunen,Ja,&Savolainen,2007) consider it amorepracticalmethod topresentDRLsas afunctionofpatientprojectionthickness.WhenpaediatricDRLsarepresentedasacurve,hospitalscancomparetheirpatientdosesdirectlyagainstthegraph,andtheneed for a large number of patients is significantly reduced. Some authorsconsidered this method more adequate than the one established by the UKNational Radiological Protection Board (NRPB-UK, now Public Health England) forsettingpaediatricDRLs(Hart,Wall,Shrimpton,Bungay,&Dance,2000).

Attentivetothechangeofparadigminmedicalimagingbroughtaboutbytheshiftto DR systems, the EC became conscious of the need to revise the existingguidelines and published an invitation to tender for a service contract, regardingEuropean Guidelines on Diagnostic Reference Levels for Paediatric Imaging(EuropeanCommission,2013b).

The27-monthprojectentitledEuropeanDiagnosticReferenceLevelsforPaediatricImaging (PiDRL)wasawarded toa consortium (ESR, EFRS, ESPR, EFOMP,&STUK,2013)composedbythe:

• EuropeanSocietyofRadiology(ESR);• EuropeanFederationofRadiographerSocieties(EFRS);• EuropeanSocietyofPaediatricRadiology(ESPR);• EuropeanFederationofOrganisationsforMedicalPhysics(EFOMP);• FinnishRadiationandNuclearSafetyAuthority(STUK).

TheconsortiumissupportedbyanExpertAdvisoryPanelformedbyrepresentativesoftheIAEA,theWHO,theCardiovascularandInterventionalRadiologicalSocietyofEurope(CIRSE),theICRP,thePublicHealthEngland(PHE,formerlyHPA)aswellasbyanexpertfromtheUnitedStatesofAmerica.



64

TheaimsofthePiDRLprojectGuidelinesare:

• torecommendamethodologyforestablishingandusingDRLsforpaediatricradiodiagnosticimagingandIRpractices;

• to update and extend the European DRLs for these examinations wheresufficientexperienceanddataareavailableforaconsensusonDRLvalues;

• topromotetheestablishmentanduseofDRLsinpaediatricradiodiagnosticimaging and IR practices so as to advance optimisation of radiationprotectionofpaediatricpatients.

DRLsareagood tool tocomplywith theALARAprincipleas theycanprovide thestimulustomonitorandpromoteimprovementsinpatientprotection,byincreasingdoseawarenessandfocusingpaediatricpracticeonachievingtherequiredimagingqualitythatpatientsneed.

Despite this scientific and legally binding framework and the recommendationsfromtheinternationalorganisations,theuseofpaediatricDRLs indailypracticeisfarfrombeingarealityintheEU.

Arecentreport fromtheDoseDataMed2 (DDM2)project (EuropeanCommission,2014b) shows thatmost EU countries have never establishedDRLs for paediatricimaging and that those few countries that have, just copied the recommendedvaluesfromtheEUguidancedocumentsthat,asalreadymentioned,entailseverallimitations.OnlyaboutonefifthofthecountriesdefinedtheirDRLsbasedontheirownnationalpatientdosesurveys(figure10).

Figure10:DRLsforpaediatricplainradiographyinEuropeancountries

FromDDM2(EuropeanCommission,2014b)

TheDDM2 report clearly shows thatdata collectionandanalysis is difficult at EUand even national levels, due to the lack of harmonisation of radiographic



65

procedures and patient categorisation. The review of current DRL systems hasshown that there is an insufficient recording of the procedures used to establishDRLs and that the available information also reveals large differences inapproaches.There isa lackofconsistencyregardingpatientgrouping(age,weightorothergroupswithavarietyofoptions)andlackofclearrecommendationsonthedosequantitiestobeused.Thediscrepanciesfoundamongstexamfrequenciesarealsosuggestiveoftheexistenceoflimitationsonthedatainfrastructure.

There is an urgent need to develop a European coding system for radiologicalprocedures,bothforadultandpaediatricpatients,tobeusedbyallmemberstatesas a tool to easily fulfil the requirement of article 64 of Directive2013/59/EURATOM,regardingtheestimationofpopulationdose:“MemberStatesshallensurethatthedistributionofindividualdoseestimatesfrommedicalexposureforradiodiagnosticandinterventionalradiologypurposesisdetermined,takingintoconsideration where appropriate the distribution by age and gender of theexposed.”

This would be a fundamental tool for future population dose studies and wouldaddressoneoftheneedsidentifiedintheDDM2report:"inordertocompareX-rayexaminationfrequencydatabetweencountries,andtoassigntypicaleffectivedosevalues to examinations, it is crucial that an “X-ray examination” is defined andcountedinaconsistentway”.

AEuropeancodingsystemforradiologicalprocedureswouldalsocontributetotheharmonisationof the“language” formedical imaging and therapy across Europe,giving healthcare providers a powerful tool for the future planning of healthsystemsatlocal,regional,nationalandEuropeanlevels.Suchaprojectshouldalsoinclude the acquisition of data on the long-term consequences of radiationexposure.



66

§



67

1.8 ThePortuguesecontextofDiagnosticReferenceLevels

The Portuguese health authorities have never promoted the establishment ofclinical audit, education and training in radiation protection, the estimation ofpopulationdose,andDiagnosticReferenceLevels,asdefinedinthe97/43EuratomDirective(EuropeanCommission,1997).

FortheEuropeanDOSEDATAMEDIIproject(EuropeanCommission,2014b;Telesetal., 2013) a groupofPortuguese researchers, scientific andprofessional societies,withthesupportofsomeregionalhealthauthorities,underthecoordinationoftheNuclear Technological Institute (ITN), joined efforts in order to estimate thePortuguese population dose using an international methodology (EuropeanCommission,2008b).

It is important to point out that until today, Portuguese health authorities haveneverdefinedDRLs,neitherforpaediatricnorforadultimaging.

Nevertheless some researchers have conducted a national study to establishnational computed tomography DRLs for adults and the paediatric population(Santos et al., 2014). The study (Santos et al., 2014) calculated and proposedPortuguese DRLs for adults and the paediatric population (Table 2 and 3). Largevariations in volume computed tomography dose index (CTDIvol) and dose-lengthproduct(DLP)valueswerefoundbetweenclinicalsites.

Table2:ProposedPortugueseCTDRLsforadultMSCTexaminationsdescribedasCTDIvolandDLP

values.

Table3:Proposedage-categorisednationalpaediatricCTDRLsdescribedasCTDIvolandDLP

values.



68

The results from this study indicate that themost common adult Portuguese CTexaminations are head, chest and abdomen, with DRLs of 75, 14 and 18 mGyCTDIvol, respectively. These values are approximately 30% higher than Europeanrecommendations.Differencesinprotocolstoimage-specificanatomicalregionsarenon-standardised, and a 20% dose variationwas found across centres nationally.DRLs have also been established for six other CT examinations. The majority ofproposedadultDRLsforCTexaminationsarehigherthanthoseofotherEuropeanCTDRLs.

For paediatric head and chest CT examinations, the proposed DRLs described asCTDIvolare48and43mGycmfornew-borns,50and6mGyfor5-yearoldchildren,70 and 6 mGy for 10-year olds and 72 and 7 mGy for 15-year old children,respectively.

ForpaediatricheadandchestCTexaminations,theproposedDRLsdescribedasDLPare 630 and 43 mGy.cm for new-borns, 767 and 139 mGy.cm for 5-year oldchildren,1096and186mGy.cmfor10-yearoldsand1120and195mGy.cmfor15-yearoldchildren,respectively.

The proposed DRLs for paediatric head CT examinations are higher than theEuropeanvalues,whereas theproposedchestCTexaminationDRLsare similar toEuropeanvaluesacrossallagecategories.

Someother studieswerecarriedout in interventional cardiology (GracianoPaulo,Rocha,Lavandeira,Costa,&Marques,2010;GracianoPaulo&Santos,2012b)asafirst attempt to analyse exposure to ionising radiation in adult patients and staffduring interventional cardiac procedures. These studies have identified an urgentneed to implementoptimisationprogrammes in interventional cardiology.Due tothe importance of the results obtained and the optimisation programmeimplementedat local level, thestudiescarriedoutsince2010wereawardedwiththe national prize (1st place, 2012) of the Associação Portuguesa para oDesenvolvimentoHospitalar(APDH)(GracianoPaulo&Santos,2012a).

Considering the results of the referred studies and the lack of education andtraining in radiation protection, the Associação Portuguesa de Intervenção emCardiologia (APIC), has implemented a continuous professional developmentprogramme in radiation protection and optimisation in interventional cardiology.Theprogrammeaims to raiseawarenessamongsthealthprofessionalsworking incath-labsabout thedangersarising from ionising radiationand the importanceofoptimisingproceduresaccordingtotheALARAprinciple.

Unfortunatelynostudyhasbeencarriedout inPortugal todeterminenationaloreven local DRLs for interventional cardiology procedures in paediatric patients,consideredmorecomplexthaninadultsandthereforeresultinhighpatientdoses(Ubeda, Vano,Miranda, & Leyton, 2012). The publications from literature in this



69

field are very scarce. Themost recently published studywith the highest sampleestablishing localDRLs (Ubeda,Miranda,&Vano,2015) foundareasonable linearcorrelation between KAP and bodyweight and proposed DRLs by patientweightgroups(table4).Theresultsfromthisstudyareagoodbasisforfutureresearchinthisarea.

Table4:KAPP75values(Gy.cm2)fordiagnosticpaediatricinterventionalcardiologyprocedures

Weight'group'(kg)

KAPP75'(Gy.cm2)

<10 0.82

10'('<20 2.45

20'(<30 4.08

30'(<40 5.71

40'(<50 7.34

50'(<60 8.97

60'(<70 10.6

70'(<80 12.2

It is essential that Portuguese health authorities and health professionalsunderstand the importance of defining DRLs, not only because of the legalrequirement, but because it is important for the quality of care delivered to thepatientsaswellasfortheprotectionofstaff.



70

§



71

2 EstablishmentofDRLsinpaediatricplainradiographyConsidering the absence of guidelines, actions and recommendations by thePortuguesehealthauthoritiesregardingtheestablishmentofDRLs,theonlyofficialreferencetouseasastartingpointforthisstudy,wastheEUguidancedocumentRP109(EuropeanCommission,1999)whichrecommendsthatDRLsfordiagnosticradiologyshouldbebasedondosesmeasuredinvarioustypesofhospitals,clinicsandpracticesandnotonlyinwell-equippedhospitals.

AccordingtotheRP109guidancedocument,theDRLinplainradiographyshouldbehigherthanthemedianormeanvalueofthemeasuredpatientdosesordosesinaphantom. Knowing that the curve giving the number of examinations and theirdoses is usually skewed with a long tail, it is recommended to use the 75thpercentile(P75)valueasareference.Theuseofthispercentileisapragmaticfirstapproach to identifying those situations in most urgent need of investigation(EuropeanCommission,1999).

Following those recommendations, the 75th percentile of KAP and ESAK wascalculated from the data collected prospectively in Hospitals A, B and C. Asdescribedinsection1.1,thesethreehealthcareinstitutionsaretheonlyreferencehospitalsforpaediatricpatientsinPortugalandarethereforerepresentativeofthepaediatric population, with practitioners exclusively dedicated to paediatricspathologiesandequippedwithup-to-datetechnology.



72

§



73

2.1 Materials and methods to determine national DRLs for chest,abdomenandpelvisplainradiography

During 2012 and 2013, anthropometric data (weight, height of patient andthicknessoftheirradiatedanatomy)wascollectedfrom9,935patients,referredfora radiography procedure to one of the three dedicated hospitals for children inPortugal(HospitalsA,BandC).Examsmadewithmobileunitswerenotincludedinthisstudy.InstitutionalReviewBoardandethicalcommitteeapprovalwasobtainedforthisstudy.

NationalDRLswerecalculatedforthethreemostfrequentX-rayproceduresmadeat hospitals A, B and C: chest Antero-Posterior (AP)/Postero-Anterior (PA)projection; abdomen AP projection; pelvis AP projection. Exposure factors andpatient dose were collected prospectively at the clinical sites. All exams werevalidatedbyboth the radiographerandtheradiologistandconsideredacceptablefortheclinicaltask.

Data collectionwas carried outwithout interferingwith the technical options foreachimagingprocedureusedineachdepartment.

Patientanthropometriccharacteristics (weight,heightandBodyMass Index -BMI)wereanalysedineachagegroupasaprocesstodefineastandardpatientforeachofthegroupsandtoallowfuturecomparisons.

Theexposureparameters(tubevoltage(kV);tubecurrent-timeproduct(mAs)andexposure time (ExT-ms), Source Skin Distance (SSD-cm), as well as kerma-areaproduct (KAP-mGy.cm2) and entrance surface air kerma, including backscatter(ESAK-µGy) were also recorded (International Commission on Radiation Units &Measurements,2005).

KAP and ESAK values were collected directly from the equipment console andmanuallyregisteredinatable,togetherwithpatientdata,includingexaminationIDnumber,toallowfutureanalysis,ifnecessary.

All devices measuring KAP (Diamentor M4-KDK, PTW®, Germany) had a validmanufacture calibration certificate, in accordance with IEC 60580 (InternationalElectrotechnicalCommission,2000),withanaccuracyof±5%andweredesignedtomeasure KAP according to IEC 61267 (International Electrotechnical Commission,2005a).

Althoughtheradiographyequipmenthadqualitycontrolmaintenanceprovidedbythemanufacturer, equipment constancyusing a calibratedRaySafe® XI dosimetrysystem(Sweden;www.raysafe.com)wasalsotested.



74

The measurements of weight, height and anatomical structure thickness of thepatients weremade using the same devices. A paediatricmeasurement rod wasused to measure the anteroposterior thickness: from the dorsal region to themiddleofthesternum(forchest);fromthelumbarregiontotheumbilicus(fortheabdomen);fromthesacrumregiontothesymphysispubis(forpelvis).

HospitalAusesaS/FsystemandisequippedwithaDiagnostfloorstandstandard(FS-S)Bucky tableplus a vertical stand (VS) andaRO1750X-ray tubeassembledwithadigitalgeneratorOptimus50kW(PhilipsHealthcare®,TheNetherlands).

HospitalBusesaDRsystemEssentaDRcompactwitha flat-panel-detectorbasedonamorphoussiliconwithagadoliniumoxysulfidescintillatorandaRO1750X-raytubeassembledwithadigital generatorOptimus50kW (PhilipsHealthcare®,TheNetherlands).

HospitalCusesaDRsystemDefinium6000ceiling-suspendedDRsystemwithaflat-panel-detector basedon amorphous siliconwith a caesium-iodide scintillator andan overhead tube assembly, model 5139720, with a digital generator of 65 kW(GeneralElectric®,USA).

Despite the fact that there are only three hospitals exclusively dedicated topaediatric patients at national level, there are no guidelines or standards ofpractice, with recommendations on what exposure parameters and technicalfeaturestouseforplainradiography.



75

2.2 Results of national DRLs for chest, abdomen and pelvis plainradiography

The anthropometric characteristics of 9,935 paediatric patients referred to aradiography procedure to one of the three dedicated paediatric hospitals inPortugalwereevaluated.

Patientagevariedfromnewbornto18years.50.4%(n=5005)ofpatientsweremaleand49.6%(n=4930)female(figure11).

Figure11:Patientdistributionbygender

Table5summarisestheanthropometriccharacteristicsofthepatients.Takingintoconsiderationthatdatadistributionisnotsymmetric,themedianvaluewasusedasthepreferredmeasureofcentraltendencytoprovidethenecessaryinformationtodefineastandardpatientforeachagegroup,asthemedianvalueislessaffectedbyoutliersandextremevalues.

Table5:PaediatricpatientsweightheightandBMI(byagegroups)

Themedianvaluesofpatientweight,heightandBMIineachagecategoryare;(<1):7kg,66cmand16kg/m2;(1to<5):14kg,93cmand16kg/m2;(5to<10):26kg,125cmand16kg/m2;(10to<16):48kg,154cmand20kg/m2;(16to18):58kg,166cmand21kg/m2,respectively.



76

Ofparticularnoteisthehighdiversityofweight,heightandBMIvalueswithineachagegroup,especially inagegroup10to<16years,withahighnumberofoutliersandextremevaluesasshowninfigures12,13and14.

Figure12:Weightperagegroupboxplot

25%-75%;-median;I - areawithoutoutliers;o–outliers;*-extremeoutliers

Figure13:Heightperagegroupboxplot




77

Figure14:BMI(kg/m2)peragegroupboxplot


Thishighdatadispersionofpatientanthropometriccharacteristicsshowninfigures12 to 14, evidences the challenges radiographers have to face in daily clinicalpractice when trying to choose the most adequate exposure parameters inpaediatricradiologyinaharmonisedandoptimisedway.

Table 6 summarises another important patient characteristic to take intoconsiderationwhenperformingaplainradiography:thethicknessoftheanatomicalstructure at the central point of exposure. Awide range of thickness valueswasfoundforeachanatomicalstructureineachagegroup.

Table6:Chest,abdomenandpelvisthicknessperagegroup

n median P75 (min-max) n median P75 (min-max) n median P75 (min-max)

<1 320 11 12 7-16 95 12 12 8-13 230 8 9 6-16

1-<5 673 13 14 10-24 120 14 15 10-18 259 10 12 7-17

5-<10 416 15 16 11-21 112 15 16 11-19 213 13 14 9-20

10-<16 326 18 19 13-33 105 17 19 12-25 214 16 18 10-27

16-≤18 254 20 21 15-24 98 20 27 15-27 194 19 20 15-28

age groups (years)

Chest (thickness, cm) Abdomen (thickness, cm) Pelvis (thickness, cm)

Themedianvaluesofpatientchest,abdomenandpelvisthickness,measuredinoursample, are similar to those published in literature (Hart et al., 2000;Wambani,Korir,Korir,&Kilaha,2013).



78

Table 7, 8 and 9 shows the exposure parameters (kV, mAs, ExT), the use ofAutomaticExposureControl(AEC)andanti-scattergrid,ineachhospitalpractice,toperform an AP/PA chest, an abdomen A/P and a pelvis A/P plain radiography,respectively. The percentage shown in AEC and grid column is related to thefrequencyoftheusage(0%never;100%always).

Table7:ExposureparametersofchestAP/PAprojection

Withregardtotable7itisimportanttonotethattheradiographersfromHospitalsAandBalwayschosetouseananti-scattergridandneverusedAEC,regardlessofthepatients’age.InHospitalAthemeanvaluesofkVusedtoperformachestplainradiography are significantly lower and themean values of mAs are significantlyhigherinallagegroups(p<0.05,ANOVAstatisticaltestwithPostHocTestStudent-Newman-Keuls-SNK).HospitalCshowssignificantlyhighermeanvaluesofExTinallagegroups (p<0.05,ANOVAstatistical testwithPostHocTest - SNK). There isnosignificantdifferencebetweentheSSDvaluesusedatthethreehospitals.



79

Table8:ExposureparametersofabdomenAPprojection

Asshownintable8,theradiographersfromHospitalsAandBalmostalwaysusedananti-scattergridwhenperforminganabdomenprocedure,independentlyofthepatients’age.AECisusedwithoutlogiccriteriainthethreehospitals.InHospitalAthe mean values of kV used to perform an abdomen plain radiography aresignificantly lower in all age groups and themean values ofmAs are significantlyhigherinagegroup5-<10(p<0.05,ANOVAstatisticaltestwithPostHocTest-SNK).HospitalBevidencessignificantlylowermeanvaluesofExTinagegroups5-<10and10-<16(p<0.05,ANOVAstatisticaltestwithPostHocTest-SNK).



80

Table9:ExposureparametersofpelvisAPprojection

Asshownintable9,theradiographersfromHospitalAalwaysusedananti-scattergridwhenperformingapelvisprocedure, independentlyof thepatients’age.AECwasusedwithoutlogiccriteriainthethreehospitals.InHospitalAthemeanvaluesof kV used to perform a pelvis plain radiography are significantly lower and themean values of mAs are significantly higher in all age groups (p<0.05, ANOVAstatistical testwithPostHocTestSNK).There isnosignificantdifferencebetweenhospitalsregardingtheSSD.

Ascanbeseenfromthetables7to9,theradiographerschosedifferentexposureparameters,evenwithinthesamehospitalandforthesameagegroupofpatients.This is clearlydemonstratedby thehigh rangeof valuesof kV,mAs,ExTandSSDusedforthesameplainradiographyprocedure.Alsoadditionaltechnicalfeatures,suchasAECortheuseofananti-scattergridareuseddifferently.



81

2.2.1 NationalDRLsbyagegroups

DRLs will be proposed as the P75 value, considering paediatric patientcategorisationbyagegroups(<1,1-<5,5-<10,10-<16,16-≤18)andbyweightgroups(<5kg;5-<15kg;15-<30kg;30-<50kg;≥50kg)assuggestedinthepreliminaryreportofthePiDRLproject(Damilakis,2015).

According to the methodology described in section 2.1, the KAPP75 and ESAKP75valueswerecalculatedforchest,abdomenandpelvisandareshownintables10,11and12respectively.

Table10:KAP&ESAKvaluesforchestAP/PA(byagegroups)

P75 min max median P75 min max median

<1 13 2 140 8 145 7 380 49

1-<5 19 2 352 11 181 5 489 52

5-<10 60 2 378 20 209 5 706 98

10-<16 134 3 525 51 234 4 771 69

16-≤18 94 3 770 49 88 4 873 55

ESAK(µGy)KAP(mGy.cm2)agegroups(years)

AwidedispersionofKAPandESAKvaluesforchestplainradiography(table10)wasobserved ineachagegroup.ThehighKAPP75andESAKP75valuesofagegroup10-<16,whencomparedtotheotheragegroups,areduetoahighvariationinvalues,most likelybecause in this agegroup thepatients’ anthropometric characteristicsshowahigherdisparity.Theoppositewasfoundinagegroup16-≤18.

Table11:KAP&ESAKvaluesforabdomenAP(byagegroups)


<1 25 4 31 23 70 10 83 63

1-<5 84 10 301 35 191 19 438 69

5-<10 140 12 2031 75 198 24 2751 101

10-<16 442 27 3051 153 583 53 2688 228

16-≤18 1401 72 3481 246 1258 96 3041 428

ESAK(µGy)agegroups(years)

KAP(mGy.cm2)

AwidedispersionofKAPandESAKvaluesforabdomenplainradiography(table11)wasobservedineachagegroup,withhigherincidenceinagegroups5-<10,10-<16and16-≤18.ConsiderablyhighervaluesofKAPP75andESAKP75wereobservedinagegroup 16-≤18, most likely because of a non-adequate exposure protocol for thisprocedure.



82

Table12:KAP&ESAKvaluesforpelvisAP(byagegroups)


<1 29 2 271 14 125 4 686 80

1/<5 75 3 626 31 158 11 1272 97

5/<10 143 5 1025 69 232 10 1443 155

10/<16 585 12 3965 224 624 25 5416 244

16/≤18 839 48 2758 254 1204 79 2507 303

ESAK:(µGy)KAP:(mGy.cm2)age:groups:(years)

AwidedispersionofKAPandESAKvaluesforpelvisplainradiography(table12)wasobservedineachagegroup,withhigherincidenceinagegroups5-<10,10-<16and16-≤18. Considerably higher values of KAPP75 and ESAKP75 were observed in agegroup 16-≤18, most likely because of a non-adequate exposure protocol for thisprocedure. Forthepurposeofthisthesis,theKAPP75andESAKP75valuespresentedinthetableswill be considered as the “1st National DRLs per age group” in order to permitfurtheranalysisoftheimpactoftheoptimisationprogrammeonpatientdoses.



83

2.2.2 NationalDRLsbyweightgroups

Asalreadymentioned,themostfrequentpatientgroupingfoundinliteratureisbypatientage.However,someauthorsstatethatweightisgenerallymorerelevantasa parameter for patient grouping for DRLs in body examinations (Järvinen et al.,2015; Watson & Coakley, 2010). In our dataset, KAP was plotted against bodyweight. Both variables were found to have a statistically significant moderatepositive correlation (p<0.005):R2=0.366,R=0.605 for chest;R2=0.335,R=0.579 forabdomen;R2=0.436,R=0.66forchest(G.Paulo,Vaño,&Rodrigues,2015).

Since thiswas a prospective study, patientswereweighed at themoment of theexam. Having collected each individual patient’s weight, the weight groups weredefined following the recommendations of the preliminary report of the PiDRLproject(Damilakis,2015):<5kg;5-<15kg;15-<30kg;30-<50kg;≥50kg.

The KAPP75 and ESAKP75 values were calculated for chest, abdomen and pelvisaccordinglyandareshownintables13,14and15respectively.Asobservedintheagegroups,awidedispersionofKAPandESAKvalueswasalsofoundinallexams.

Table13:KAP&ESAKvaluesforchestAP/PA(byweightgroups)


<5 21 2 138 9 140 10 380 68

50<15 16 2 352 9 155 5 489 49

150<30 28 2 447 15 199 5 573 57

300<50 131 3 608 45 228 4 706 95

≥50 145 7 770 59 189 8 873 62

ESAK:(µGy)KAP:(mGy.cm2)weight:groups:(kg)

ItisimportanttonotethehighKAPP75valuesobservedforchestplainradiographyexams inweightgroups30-<50kgand≥50kgwhencompared to theothers (table13),mostlikelyduetoaninappropriateexposureprotocolforthisprocedure.



84

Table14:KAP&ESAKvaluesforabdomenAP(byweightgroups)


<5 30 4 128 20 70 10 83 59

5.<15 35 10 249 25 82 19 345 63

15.<30 138 12 2031 71 199 24 2751 102

30.<50 248 24 3051 96 285 16 569 134

≥50 1445 62 3481 410 1322 79 3041 479

ESAK9(µGy)weight9

groups9

(kg)

KAP9(mGy.cm2)

The high KAPP75 and ESAKP75 values observed for abdomen plain radiography inweightgroups15-<30kg,30-<50kgand≥50kg (table14)aremost likelydue toaninappropriateexposureprotocolforthisprocedure.

Table15:KAP&ESAKvaluesforpelvisAP(byweightgroups)


<5 239 5 241 9 480 28 480 71

52<15 37 2 436 21 125 4 746 84

152<30 120 3 944 68 216 10 1443 141

302<50 314 12 1540 132 327 25 2181 190

≥50 1164 31 3965 356 1217 50 5416 379

ESAK:(µGy)KAP:(mGy.cm2)weight:groups:(kg)

Unexpectedly,anabnormallyhighKAPP75andESAKP75valuewas found inpatientswithlessthan5kg.Theauthorsconsiderthatthisisrelatedtotheinadequateusageof gonads protective shields, combinedwith inappropriate use of AEC. Also highKAPP75andESAKP75valueswere found inweightgroup≥50,most likelydue toaninappropriateexposureprotocolforthisprocedure. For the purpose of this thesis, the KAPP75 and ESAKP75 values, presented in thetableswillbeconsideredas the“1stNationalDRLsperweightgroups” inorder topermit further analysis of the impact of the optimisation programme on patientdoses.



85

2.2.3 NationalversuslocalDRLs

Thedatacollected fromthe threepaediatrichospitalsof this study (phase1)andpresentedintables10to12givesafirstindicationofwhatcouldbeconsideredasthe“1stNationalDRLs”forchest,abdomenandpelvisplainradiographyinPortugal.However, looking at the high discrepancy in exposure parameters and in thetechnical featuresusedforeachexamination(table7,8and9)andgiventhefactthat one of the hospitals used an S/F system, significant differences were foundbetween KAP and ESAK values at each hospital, due to an obvious lack of aharmonisationofpractice.

ThediscrepancybetweenKAPandESAKvaluesineachhospitalisclearlyevidencedin figures15to20. Ingeneral, thehighestKAPandESAKvalueswereobserved inhospitalA thatusedaS/Fsystem.ThehighKAPandESAKvalues fromHospitalAhadanegativeimpactonthe“1stNationalDRLs”andwere3to4timeshigherthantheDRL values found in literature (Billinger,Nowotny,&Homolka, 2010; Roch&Aubert, 2013; Kristien Smans et al., 2008; Sonawane, Sunil Kumar, Singh, &Pradhan,2011;Wambanietal.,2013).

0"

100"

200"

300"

400"

500"

600"

700"

<1" 1+<5" 5+<10" 10+<16" 16+≤18"

KAP"mGy.cm2"

1st"Na:onal"

Hosp"A"

Hosp"B"

Hosp"C"

Figure15:ComparisonoftheHospitals’KAPP75valuewiththe“1stNationalDRL”forchestplain

radiography

TheKAP75valuefromHospitalAforchestplainradiographyishigherthanthe“1stNationalDRL”inallagegroups.



86

0"

100"

200"

300"

400"

500"

600"

<1" 1*<5" 5*<10" 10*<16" 16*≤18"

ESAK"µGy"

1st"Na7onal"

Hosp"A"

Hosp"B"

Hosp"C"

Figure16:ComparisonoftheHospitals’ESAKP75valuewiththe“1stNationalDRL”forchestplain

radiography

TheESAK75valuefromHospitalAforchestplainradiographyishigherthanthe“1stNationalDRL” inallagegroups.Thesame isverified inHospitalB inagegroup5-<10.

0"

500"

1000"

1500"

2000"

2500"

3000"

<1" 1(<5" 5(<10" 10(<16" 16(≤18"

KAP"mGy.cm2"

1st"Na8onal"

Hosp"A"

Hosp"B"

Hosp"C"

Figure17:ComparisonoftheHospitals’KAPP75valuewiththe“1stNationalDRL”forabdomenplain

radiography

TheKAP75valuefromHospitalAforabdomenplainradiographyishigherthanthe“1stNationalDRL”inallagegroups(wheredataisavailable).ThesameisverifiedinHospitalCinagegroup10-<16.



87

0"

500"

1000"

1500"

2000"

2500"

3000"

3500"

<1" 1(<5" 5(<10" 10(<16" 16(≤18"

ESAK"µGy"

1st"Na6onal"

Hosp"A"

Hosp"B"

Hosp"C"

Figure18:ComparisonoftheHospitals’ESAKP75valuewiththe“1stNationalDRL”forabdomenplain

radiography

TheESAK75valuefromHospitalAforabdomenplainradiographyishigherthanthe“1stNationalDRL”inallagegroups(wheredataisavailable).ThesameisverifiedinHospitalCinagegroup10-<16.

0"

500"

1000"

1500"

2000"

2500"

3000"

<1" 1(<5" 5(<10" 10(<16" 16(≤18"

KAP"mGy.cm2"

1st"Na8onal"

Hosp"A"

Hosp"B"

Hosp"C"

Figure19:ComparisonoftheHospitals’KAPP75valuewiththe“1stNationalDRL”forpelvisplain

radiography

TheKAP75valuefromHospitalAforpelvisplainradiographyishigherthanthe“1stNationalDRL”inallagegroups.ThesameisverifiedinHospitalCinagegroup16-≤18.



88

0"

500"

1000"

1500"

2000"

2500"

3000"

3500"

<1" 1(<5" 5(<10" 10(<16" 16(≤18"

ESAK"µGy"

1st"Na6onal"

Hosp"A"

Hosp"B"

Hosp"C"

Figure20:ComparisonoftheHospitalESAKP75valuewiththe“1stNationalDRL”forpelvisplain

radiography

TheESAK75valuefromHospitalAforpelvisplainradiographyishigherthanthe“1stNationalDRL”inalltheagegroups.ThesameisverifiedinHospitalBinagegroup1-<5.

TheheterogeneityofKAPP75andESAKP75valuesshowninfigures15to20forchest,abdomenandpelvisplainradiography,clearlyindicatestheusageofahighvarietyofprotocolsforthesameprocedurebydifferentradiographers(x-axis).Toevidencethat variety an example is shown in figure 21 about the choice of kV (y-axis) forchestplainradiographyineachagegroup.

Figure21:MeankVvaluesusedbyeachradiographerforchest

plainradiographyineachpatientagegroup



89

The results shown in section 2.2 were presented to the radiographers andradiologistsateachhospital.Noneofthethreehospitalshaseverimplementedanykindofoptimisationprocessorhasanalysedtheprotocolsandpre-setsconfiguredin their equipment. All radiologists and radiographers were receptive tooptimisationprocedureswiththeobjectivetoreduceexposuretopatients.



90

§



91

2.3 Limitationsofsection2

The paediatric DRLs were established from data collected at three dedicatedregional public paediatric centres, each representing geographic regions ofPortugal.However,itisimportanttotakeintoconsiderationthatalargeproportionof paediatric plain radiography procedures is performed in general hospitals,potentiallywithout tailoredpaediatricprotocols.Thisdosedatawasnotanalysedduringinthisstudy.

Bycoincidence,thePortugueseHealthMinisterandtheregionalhealthauthoritiesdecided to closeHospital Awhen the results of this surveywere presented. ThishospitalhadthelastradiologydepartmentinPortugalthatstillusedanS/Fsystem.Due to the closure, no additional optimisation measures were possible for plainradiographyproceduresforthosesystemsthatpresentedthehighestdosevaluesinthisstudy.

Due to limitedhumanand financial resources itwasonlypossible todevelop theoptimisationprocessinonehospital.Consideringtheresultsfromsection2.2,itwasconsideredthatHospitalCwouldbenefitmostfromanoptimisationprogramme.



92

§



93

3 PlainradiographyoptimisationphantomtestsThe findingsof thedata collection inphase1of this researchwork indicated theneed for further investigation with regard to optimisation of plain radiographyprotocols forpaediatricexaminations.A literaturereviewof relevantoptimisationtests in plain radiography examinations was undertaken and is presented in thischapter.

Thenextstageofthisresearchworkinvolvedexperimentaltestingtooptimisethethreemostfrequentproceduresinplainradiography.Thisresearchactivityincludedtheuseofqualityassurance(QA)andanthropomorphicphantoms.Thefindingsofthisphaseofthestudyaredescribedanddiscussedinthischapter.

3.1 Optimisationinplainradiography

The European guidelines for paediatric imaging (European Commission, 1996a)provide a baseline for optimisation in terms of minimum tube potential andmaximumESAK.Changes inpractice tomeet theseguidelinesmustbeperformedcautiously,asincreasedX-raypenetrationmayreduceimagecontrastandreduceddosecancauseunacceptablesignal-to-noiseratios(Martinetal.,2013).

Radiationdosestopaediatricpatientsfromplainradiographyarerelativelylow,butbecauseofthehighfrequencyoftheseprocedures,theiroptimisationisimportantfortheradiologypractice(UNSCEAR,2013).

Foranoptimisationstrategytobeeffective,allhealthprofessionalsinvolvedintheuse of X-ray equipment need to have knowledge and access to the results ofperformance tests and patient dose surveys. In addition there should be acontinuing programme of assessment to track any changes in equipmentperformance. There should be close links between the radiographer, themedicalphysicistandtheradiologisttoprovideagreateropportunityforoptimisation(C.J.Martin,LeHeron,Borrás,Sookpeng,&Ramirez,2013).

Themainobjectiveofoptimisationof radiologicalprocedures is toadjust imagingparameters and implement measures in such a way that the required image isobtained with the lowest possible radiation dose and maximised benefit(InternationalCommissiononRadiologicalProtection,2013).

To achieve this goal, good practice in radiographic technique is needed andtherefore special attentionmust given, simultaneously, to several aspects of theprocedure,suchas:a)patientpositioningandimmobilisation;b)accuratefieldsize



94

and correct X-ray beam limitation; c) the use of protective shielding, whenappropriateandd)optimisationofradiographicexposurefactors.

Using a correct beam limitation is crucial to avoid unnecessary radiation doseoutside the area of interest, and prejudice the image contrast and resolution byincreasing the scattered radiation. Therefore proper collimation of the examinedstructureisnecessaryandpreferredovertheuseofpost-processingtoolsinCRorDRsystems,suchasimagingcrop.Thispost-processingtoolcanpotentiallyhideanincrease in patient dose due to an unnecessarily overexposed area (InternationalCommissiononRadiologicalProtection,2013;Mooreetal.,2012).

ChoosingthemostadequateexposurefactorsandmakingthebesteffectiveuseofthetechnologicalfeaturesavailableintheX-rayequipment,iscrucialforobtainingthebestdiagnosticqualityimagewiththelowestpossibledose.

ThekV,especiallywhenusingdigitalsystems,shouldbeusedatthehighestvalue,within theoptimal range,consideringthepositionandanatomicalstructurebeingexamined, allowing the lowest quantity of mAs needed to provide an adequateexposuretotheimagereceptor(Herrmannetal.,2012).

Forchest,abdomenandpelvisexposure,theprincipleofusingahighkVtechniqueshouldbefollowed,sinceitwouldresultinlowerpatientattenuation,andthereforelowerdoseforthesamedetectorexposure.ThekVshouldalsobeincreasedineachascending age/size groupdue to the increases in tissue thickness,which requiresmorephotonpenetration(Knight,2014).

It is importanttotake intoconsiderationthatwheneverusinghigherkVvaluesorimaging thicker structures, an increase of scatter radiation is expected andtherefore an antiscatter grid should be used. The Image Gently programme(www.imagegently.org) and the American Society of Radiologic Technologists(ASRT)WhitePaper(Herrmannetal.,2012)recommendtheuseofanti-scattergridabove10-12cmthickness.

Another important technological feature to take into consideration is the use ofadditional filtration, which removes the lower energies from the X-ray spectrumandconsequently raises theaveragebeamenergy foraconstantkV.Byremovingthe low energies from the spectrum, the ESAK to the patient is reduced (Brosi,Stuessi,Verdun,Vock,&Wolf,2011).

The commonly used filters are made of copper (Cu) or aluminium (Al) or acombinationofboth.TheEuropeanGuidelinesrecommendtheuseof0.1mmCu+1mmAlor0.2mmCu+1mmAladditionalbeamfiltration forS/Fsystems,whenconsideredappropriate(EuropeanCommission,1996a).Onepotentialconsequenceofusingadditionalfiltrationisthereductionofimagequalityduetothedecreasedimagecontrast.Howeverthisdisadvantageisofminorimportanceindigitalsystems



95

because image contrast may be selectively enhanced by using appropriate postprocessingtools(Brosietal.,2011).

Once themedicalexaminationusing ionising radiationhasbeenclinically justifiedanddecided,theproceduremustbeoptimised.Thereforetheradiationdosewhichis delivered to the patient must be ALARA, but high enough for obtaining therequired diagnostic information, taking into account economic and social factors.The written protocols (guidelines) for every type of standard practice should beoptimised(EuropeanCommission,2009).

Anoptimisationprocessshouldbeunderstoodasadynamicprocessandaspartofthe clinical audit programme of the radiology department. The optimisation ofclinical protocols for paediatric patients (Kostova-Lefterova, Taseva, Hristova-Popova,&Vassileva,2015)shouldincludethestepsdescribedinfigure22.

Figure22:Optimisationofclinicalprotocolsforpaediatricimaging

By implementing this dynamic process, it is expected to reduce patient dosewithoutinterferingwithdiagnosticimagequality.Thesecretforthesuccessofthisprocess is the continuous contribution of a multidisciplinary team involvingradiographers, radiologists and medical physicists. This concept constitutes thepillars in the development and implementation of a patient safety culture in themedicalimagingdepartment.



96

§



97

3.2 Experimental tests with anthropomorphic phantoms (objectiveimageanalysis)

Toanalysetherelationshipbetweenexposurefactors,theuseoftechnicalfeaturesanddose,experimentaltestsweremadeusingtwoanthropomorphicphantoms:A)CIRSTM ATOM model 705®; 110cm of height and 19Kg of weight and B) KyotokagakuTMmodelPBU-60®;165cmofheightand50Kgofweight(figure23).

A

B

Figure23:AnthropomorphicphantomsusedinexperimentaltestsA-PhantomCIRSTMATOMmodel705;B-PhantomKyotokagakuTMmodelPBU-60

3.2.1 Methodologyofexperimentaltestswithanthropomorphicphantoms

Phantom tests were performed at Hospital C, where the optimisation process ofthis study was carried out. Dose exposure data from the phantom tests wasobtainedusingthesamemethodologyasforthedatacollectionofpatientexposuredata.

Afterdatacollection,anobjective imageanalysiswasperformedbyanalysing thevariationofthemeanvalueofthestandarddeviation(SD),measuredinfour(greencircles) Regions of Interest (ROI) in each image (figure 24). OsiriX® software(Pixmeo,Switzerland)wasusedtoperformtheROImeasurements.

The analysis of the SD values iswell recognised as the standardmethod used toreflectthedegreeofnoisewhenimagingparametersarechanged(Sunetal.,2004;Sun, Lin,Tyan,&Ng,2012).HigherSDvaluesare related toan increaseof imagenoiseandtoadecreaseofdosevalues.



98

A-ATOM705chest B-ATOM705abdomen C-ATOM705pelvis

D-PBU-60chest E-PBU-60chestabdomen

Figure24:ExampleofROIlocations,foranalyseswithOsiriX®software(AtoE)

Several acquisitions were made with different exposure factors, differentcombinationsofAECchambersettings,additionalfiltrationandtheuseofananti-scattergridforchest,abdomenandpelvisexaminations(tables16to20).KAPandESAKvalueswerealsoregisteredforeachexposure.



99

3.2.2 Resultsofphantomsexperimentaltests

Table16showstheresultsofthetestsforthechestexaminationonCIRSTMATOMmodel705®.AllexposuresweremadewithaSIDof180cm.

Table16:ExperimentaltestsforchestexaminationusingCIRSTMATOMmodel705®

Test%ID

Tube%voltage%(kV)

Tube%current%(mA)

Tube%current%time%product%

(mAs)

Ext%(ms)

ESAK%(µGy)

KAP%(mGy.cm2)

Grid Additional%Filtration

AEC%chamber

SD%Mean

1 90 339 2 7 18 5 no 0.3Cu right 425

2 90 339 2 6 21 5 no 0.2Cu right 442

3 90 323 2 5 29 7 no 0.1Cu right 447

4 90 324 3 11 26 7 yes 0.3Cu right 421

5 100 321 3 8 26 7 yes 0.3Cu right 412

6 90 324 3 9 29 7 yes 0.2Cu right 425

7 100 321 2 7 29 7 yes 0.2Cu right 418

8 90 338 4 12 31 8 no 0.3Cu central 412

9 100 323 2 6 35 9 yes 0.1Cu right 420

10 90 324 2 7 35 9 yes 0.1Cu right 424

11 90 323 4 11 35 9 no 0.2Cu central 414

12 90 337 2 5 39 10 no none right 464

13 90 324 3 9 43 11 no 0.1Cu central 427

14 90 326 2 6 53 14 yes none right 429

15 100 322 2 5 53 14 yes none right 425

16 90 319 7 23 55 14 yes 0.3Cu central 388

17 90 334 6 18 61 16 yes 0.2Cu central 390

18 90 326 3 8 65 17 no none central 443

19 90 320 5 17 78 20 yes 0.1Cu central 398

20 90 324 5 15 121 31 yes none central 408

Thetestidentification(ID)number18correspondstotheprotocolmostfrequentlyusedatHospitalC.Consideringtherecommendationsofusingadditional filtrationand anti-scatter grid for high kV procedures, it was decided together withradiologists and radiographers from the radiology department to start using theexposure conditions of test ID number 6: use additional filtration of 0.2mm ofCopperandselecttherightAEC(correspondingtotherightlung)chamber.TheAECchamberrelatedtotherightlungwaschoseninordertoavoidtheAECtoincludethedensityofthespine,mediastinumandsternum.

TheexposureconditionsaspertestIDnumber6showareductionof55%and56%onESAKandKAPvalues, respectively,whilenot interferingonmAsandexposuretime,whencomparedwiththeonesusedatthedepartment.



100

Despite the significant dose reduction (one sample T-test, p<0.05), no differencewasfoundbetweenthevariationsofthemeanofSDvaluesobtainedfromthefourROIs(onesampleT-test,p>0.05).

Table17showstheresultsofthetestsforabdomenexaminationonCIRSTMATOMmodel705®.AllexposuresweremadewithaSIDof110cm.

Table17:ExperimentaltestsforabdomenexaminationusingCIRSTMATOMmodel705®

Test%ID

Tube%voltage%(kV)

Tube%current%(mA)


(mAs)

Ext%(ms)

ESAK%(µGy)

KAP%(mGy.cm2)

GridAdditional%Filtration

AEC%chamber SD%Mean

1 75 253 2 9 66 16 yes 0.3Cu central 307

2 70 255 4 16 70 17 no 0.3Cu central 331

3 75 252 3 8 79 19 yes 0.2Cu central 314

4 70 255 3 13 87 21 no 0.2Cu central 323

5 75 253 2 7 106 25 yes 0.1Cu central 322

6 70 255 3 11 115 28 no 0.1Cu central 325

7 80 248 5 21 153 37 yes 0.3Cu central 400

8 80 253 1 6 164 39 no none central 318

9 80 249 5 19 183 44 yes 0.2Cu central 397

10 70 255 2 9 184 44 no none central 318

11 70 250 11 46 202 48 yes 0.3Cu central 405

12 70 250 9 37 239 57 yes 0.2Cu central 385

13 80 248 4 16 240 58 yes 0.1Cu central 392

14 70 250 8 30 314 75 yes 0.1Cu central 382

15 80 252 4 14 396 95 yes none central 383

16 70 251 6 24 516 124 yes none central 377

ThetestIDnumber16correspondstotheprotocolmostfrequentlyusedatHospitalC.Considering the recommendationsofusingadditional filtrationandanti-scattergrid forhighkVproceduresor toexamineanatomicalstructureswithhighatomicnumber, it was decided together with radiologists and radiographers from theradiologydepartment to startusing theexposureconditionsof test IDnumber3:useadditionalfiltrationof0.2mmofCopperandselectthecentralAECchamber.

The exposure conditions, as per test ID number 3, shows a reduction of 85% onESAKandKAPvaluesandareductionof67%oftheExT,whencomparedwiththeonesusedatthedepartment.

Despite the significant dose reduction (one sample T-test, p<0.05), no differencewasfoundbetweenthevariationsofthemeanofSDvaluesobtainedfromthefourROI’s(onesampleT-test,p>0.05).



101

Table 18 shows the results of the tests for pelvis examination on CIRSTM ATOMmodel705®.AllexposuresweremadewithaSIDof110cm.

Table18:ExperimentaltestsforpelvisexaminationusingCIRSTMATOMmodel705®

Test%ID

Tube%voltage%(kV)

Tube%current%(mA)


(mAs)

Ext%(ms)

ESAK%(µGy)

KAP%(mGy.cm2)

GridAdditional%Filtration

AEC%chamber

Gonads%protection SD%Mean

1 75 200 4 21 129 24 yes 0.3Cu central no 386

2 75 200 5 23 138 25 yes 0.3Cu central yes 322


4 75 205 4 18 150 27 yes 0.2Cu central no 376


6 75 204 3 14 186 34 yes 0.1Cu central no 365


8 75 204 2 12 283 52 yes none central yes 282


10 75 204 2 12 293 54 yes none central no 355


12 75 200 4 22 525 96 yes none central yes 252

ThetestIDnumber12correspondstotheprotocolmostfrequentlyusedatHospitalC.Considering the recommendationsofusingadditional filtrationandanti-scattergrid forhighkVproceduresor toexamineanatomicalstructureswithhighatomicnumber, it was decided together with radiologists and radiographers from theradiologydepartment to startusing theexposureconditionsof test IDnumber4:useadditionalfiltrationof0.2mmofCopperandselectthecentralAECchamber.Itis important to note that the use of gonads protection increases ESAK and KAPvaluesandthereforeitwasrecommendednottousegonadsprotection,neitherinfemales, due to the variety of the location of ovaries, neither in males, if theprotection falls in the exposure area. The same recommendation was found inliterature(Bardo,Black,Schenk,&Zaritzky,2009;Fawcett&Barter,2009)

Theexposureconditions,aspertestIDnumber4,showareductionof71%onESAKandKAPvalues,whencomparedwiththeonesusedatthedepartment.


Table 19 shows the results of the tests for chest examinationonKyoto kagakuTMmodelPBU-60.AllexposuresweremadewithaSIDof180cm.



102

Table19:ExperimentaltestsforchestexaminationusingKyotokagakuTMmodelPBU-60

Test%ID

Tube%voltage%(kV)

Tube%current%(mA)


(mAs)

Ext%(ms)

ESAK%(µGy)

KAP%(mGy.cm2)


AEC%chamber

SD%Mean

1 100 320 3 9 32 24 yes 0.3Cu right 529

2 110 317 2 7 33 24 yes 0.3Cu right 526

3 100 320 3 8 35 26 yes 0.2Cu right 533

4 110 317 2 6 37 27 yes 0.2Cu right 528

5 100 321 4 12 41 30 yes 0.3Cu left 529

6 100 322 1 4 42 31 no none right 542

7 100 321 2 7 43 32 yes 0.1Cu right 533

8 100 319 4 13 44 32 yes 0.3Cu all 528

9 110 317 2 6 45 33 yes 0.1Cu right 530

10 100 339 3 10 46 34 yes 0.2Cu left 526

11 100 320 4 11 48 36 yes 0.2Cu all 527

12 100 338 3 9 56 41 yes 0.1Cu left 532

13 100 322 3 9 59 44 yes 0.1Cu all 528

14 100 321 2 6 66 49 yes none right 530

15 110 316 2 5 66 49 yes none right 532

16 100 334 3 8 86 63 yes none left 528

17 100 320 3 9 90 66 yes none all 531

18 100 317 9 29 99 73 yes 0.3Cu central 518

19 100 315 8 25 109 80 yes 0.2Cu central 522

20 100 316 7 22 136 100 yes 0.1Cu central 524

21 100 315 6 21 211 156 yes none central 527

ThetestIDnumber21correspondstotheprotocolmostfrequentlyusedatHospitalC.Considering the recommendationsofusingadditional filtrationandanti-scattergrid for high kV procedures, it was decided together with radiologists andradiographersfromtheradiologydepartmenttostartusingtheexposureconditionsof test ID number 3: use additional filtration of 0.2mmof Copper and select therightAECchamberrelatedtotherightlungwaschoseninordertoavoidtheAECtoincludethedensityofthespine,mediastinumandsternum.

Theexposureconditions,aspertestIDnumber3,showareductionof83%onESAKandKAPvalues,whencomparedwiththeonesusedatthedepartment.




103

Table 20 shows the results of the tests for abdomen examination on KyotokagakuTMmodelPBU-60.AllexposuresweremadewithaSIDof110cm.

Table20:ExperimentaltestsforabdomenexaminationusingKyotokagakuTMmodelPBU-60

Test%ID

Tube%voltage%(kV)

Tube%current%(mA)


(mAs)

Ext%(ms)

ESAK%(µGy)

KAP%(mGy.cm2)


AEC%chamber

SD%Mean

1 80 252 9 36 310 165 yes 0.3Cu all 354

2 90 245 6 26 310 169 yes 0.3Cu central 343

3 80 249 7 29 330 179 yes 0.2Cu all 367

4 70 250 17 70 360 195 yes 0.3Cu all 345

5 90 248 4 15 360 196 yes 0.1Cu all 360

6 90 254 6 22 370 198 yes 0.2Cu central 343

7 80 249 12 47 390 211 yes 0.3Cu central 336

8 80 248 6 23 400 217 yes 0.1Cu all 350

9 70 250 14 57 420 230 yes 0.2Cu all 344

10 80 249 10 39 450 241 yes 0.2Cu central 331

11 90 245 5 20 460 248 yes 0.1Cu central 346

12 70 250 25 99 510 276 yes 0.3Cu central 340

13 70 250 11 44 530 286 yes 0.1Cu all 341

14 90 248 3 13 540 292 yes none all 359

15 80 248 8 32 540 294 yes 0.1Cu central 343

16 70 250 20 81 600 327 yes 0.2Cu central 340

17 80 248 5 19 630 339 yes none all 346

18 90 244 4 17 700 378 yes none central 342

19 70 251 16 64 780 419 yes 0.1Cu central 340

20 80 248 7 27 860 465 yes none central 345

21 70 250 9 35 870 468 yes none all 346

22 70 250 13 52 1270 687 yes none central 338

ThetestIDnumber21correspondstotheprotocolmostfrequentlyusedatHospitalC.Considering the recommendationsofusingadditional filtrationandanti-scattergrid forhighkVproceduresor toexamineanatomicalstructureswithhighatomicnumber, it was decided together with radiologists and radiographers from theradiologydepartment to startusing theexposureconditionsof test IDnumber3:useadditionalfiltrationof0.2mmofCopperandselectalltheAECchambers.

Theexposureconditions,aspertestIDnumber3,showareductionof63%onESAKandKAPvaluesandareductionof44%oftheExT,whencomparedwiththeonesusedatthedepartment.



104


An example of the influence of the AEC chamber in plain radiography is given infigure25.ImagesAandBarefromthesamepatientandwereobtainedatHospitalC(withanintervalof2months).ChestplainradiographyAwasmadewithAECandcentral chamber (exposure conditions used at the department). Chest plainradiography B was made with AEC and lateral right chamber. The exposureconditionswerethesame(90kV;2mAs)forbothexposuresandtheKAPandESAKvalueswere3 and4 times lower respectively in chestplain radiographyB.Whenshownblindlytofourradiologists,allofthemdefinedchestplainradiographyBastheonewiththebetterdiagnosticimage.

A

B

Figure25:A:ChestplainradiographywithAEC+centralchamber;B:ChestplainradiographywithAEC+lateralrightchamber



105

3.3 Optimisedexposurecriteriaforchest,abdomenandpelvisplainradiography

The results obtained from the anthropomorphic phantoms were discussed in ameetingwith the radiologistsand radiographers fromHospitalCanda consensuswasobtainedtoharmonisepractice.

Thenewexposurecriteriaforabdomenplainradiographyweredefinedtakingintoconsideration the results of the anthropomorphic phantoms tests, therecommendations from literature and the outputs from the meetings withradiographersandradiologistsfromtheradiologydepartment.

Thenewexposurecriteriaforeachagegroupweredefinedaccordingtotheresultsobtainedfromtheanthropomorphicphantomstestsandbyreviewingtheexposurecriteriapublishedintheliterature(Amaral,Matela,Pereira,&Palha,2008;Cooketal.,2001;Donetal.,2013;Doyle,Gentle,&Martin,2005;Knight,2014;McCarty,Waugh,McCallum,Montgomery,&Aszkenasy,2001;Mooreetal.,2012;GracianoPaulo et al., 2011; Rizzi et al., 2014; K Smans, Struelens, Smet, Bosmans, &Vanhavere, 2010; Suliman& Elawed, 2013; Zhang, Liu, Niu,& Liu, 2013) and theoutcome of several group meetings held with radiographers and radiologistsworkingattheradiologydepartment.

Thenewexposurecriteria forplainradiographyofchest,abdomenandpelvisperagegrouparedefinedintables21to23.

Table21:Newexposurecriteriaforchestplainradiography

Age$groups$(years)

Median$weight$(Kg)

Tube$tension$(kV)

Chamber GridAdditional$filtration

<1 7 70 central no 0.1mm/Cu

12<5 14 80 central no 0.1mm/Cu

52<10 26 90Lateral/right/(right/lung)

yes 0.2mm/Cu

102<16 46 100Lateral/right/(right/lung)

yes 0.2mm/Cu

162≤18 58 110Lateral/right/(right/lung)

yes 0.2mm/Cu

Chest



106

For paediatric patients under 5 years, the most adequate option is to use thecentral ionisation chamber of theAEC system (considering that the patient chestdoesnotcoverlateralchambers),combinedwithanadditionalfiltrationof0.1mmofcopper,withouttheanti-scattergrid.

For patients above 5 years, themost adequate option is to use the lateral rightionisationchamberoftheAEC(correspondingtothepatient’srightlung),toavoidthe influence of the spine, the mediastinum structures (mainly the heart) andsternum, combined with an additional filtration of 0.2mm of copper, with anti-scattergrid.

Themedianofpatientweightforeachagegroupwasprovidedasanindicatorforradiographerstoadapttheexposureconditionsaccordinglyifnecessary.

Table22:Newexposurecriteriaforabdomenplainradiography

Age$groups$(years)

Median$weight$(Kg)

Tube$tension$(kV)

Chamber Grid Additional$filtration

<1 7 65 central no 0.1mm1Cu

14<5 14 70 central no 0.1mm1Cu

54<10 26 75 central yes 0.2mm1Cu

104<16 46 80 All yes 0.2mm1Cu

164≤18 58 90 All yes 0.2mm1Cu

Abdomen

For paediatric patients under 5 years, the most adequate option is to use thecentral ionisation chamber of the AEC system (considering that the patientabdomendoesnotcover lateralchambers),combinedwithanadditionalfiltrationof0.1mmofcopper,withouttheanti-scattergrid.

Forpatientsabove5years,themostadequateoptionistouseallthreechambersof theAEC, combinedwithanadditional filtrationof0.2mmof copper,withanti-scattergrid.




107

Table23:Newexposurecriteriaforpelvisplainradiography

Age$groups$(years)

Median$weight$(Kg)

Tube$tension$(kV)

Chamber Grid Additional$filtration gonads$protection

<1 7 65 central no 0.1mm1Cu

14<5 14 70 central no 0.1mm1Cu



164≤18 58 90 central yes 0.2mm1Cu

only$in$males$if$protection$is$out$of$exposure$field

Pelvis

The new exposure criteria for pelvis plain radiography were defined taking intoconsideration the results of the anthropomorphic phantom tests, therecommendations from literature and the outcome of the meetings held withradiographersandradiologistsworkingattheradiologydepartment.

For paediatric patients under 5 years, the most adequate option is to use thecentralionisationchamberoftheAECsystem(consideringthatthepatient’spelvisdoesnotcoverlateralchambers),combinedwithanadditionalfiltrationof0.1mmofcopper,withouttheanti-scattergrid.

For patients above 5 years, the most adequate option is to use the centralionisationchamberoftheAEC,combinedwithanadditionalfiltrationof0.2mmofcopper,withanti-scattergrid.

As indicatedabove, theuseof gonadsprotection increasesESAKandKAPvalues.Therefore itwas recommendednot tousegonads,neither in females,due to thevarietyofthelocationofovaries,norinmales,iftheprotectionfallsintheexposurearea.




108

§



109

3.4 Subjectiveanalysisofimagequality(methodologyandresults)

Dataforchest,abdomenandpelvisexaminations(n=24foreach)pre/postthenewexposure criteria were organised in Digital Imaging and Communications inMedicine(DICOM)formatforthefouragecategories(<1;1-<5;5-<10;10-<16)andpresented for image quality review. In total 72 cases were prepared for imageevaluation.PatientimageswerepresentedinViewerforDigitalEvaluationofX-rayimages(ViewDEX2.0)software(Hakanssonetal.,2010).

VisualGradingCharacteristic(VGC)imagequalityevaluationwasperformedblindlyby four paediatric radiologists, eachwith aminimum of 10 years of professionalexperience, using anatomical criteria scoring. The anatomical criteria consideredmost important foreachexaminationwerechosenonaconsensusbasisbetweenthefourradiologistsfromtheEuropeanGuidelinesonQualityCriteriaforDiagnosticRadiographicImagesinPaediatrics(EuropeanCommission,1996a),(table24).

Table24:Imageanalysesusinganatomicalcriteriascoringandthefivepointscale

Imageswereanalysedonadiagnosticworkstationandtheambientluminancewasmeasuredwith a RaysafeTM Xi Light detector (Unfors RaySafe, Sweden) to ensureconsistentambientlightingconditionsoflessthan40lux(Brennanetal.,2007).

ThroughVGCanalysis, theobserver usesmultiple scale steps to statehis opinionabout the image quality (M Båth&Månsson, 2007). The observers of this study



110

gavetheiropinionbyusingascalefrom1to5toclassifythefulfilmentofaspecificcriterion.Twodatasetswerethereforeobtainedbeforeandafterthenewexposurecriteria. The data from the two data sets were used to calculate the VGC datapoints.ThereferredpointsrepresenttheVGCcurvecoordinatesofaplot.Theareaunderthecurve(AUCVGC)canbetakenasameasureofthedifferenceintheimagequality between the pre and post settings. A curve equivalent to an AUCVGC ofaround 0.5 indicates equality between both settings (Ludewig, Richter, & Frame,2010). Values lower than 0.5 indicate better image quality of the “pre” settings.Valueshigherthan0.5indicatebetterimagequalityofthe“post”settings.

To perform VGC analyses, a web-based calculator for Receiver OperatingCharacteristic (ROC) Curves (John Eng., Maryland, USA) was used to analyse theAUCVGCpercriterionandradiologist (Eng,2013).VGCanalysis reviews thedataasordinalwithnoassumptionsontheAUCVGCdistribution.A0.5AUCVGCindicatesnodifferenceintheradiologists’ratingbeforeandafterthenewexposurecriteria(thenew exposure conditions were considered the true positive fraction for all theanalysis)(MBåth&Månsson,2007).

The Student’s T-test (for Independent samples)was also used to compare imageevaluation findings emanating from chest, abdomen and pelvis examinations preandpostthenewexposurecriteriaandCohen’sKappatestingtoanalysetheinterobserveragreement (p<0.05)betweenradiologistevaluations,acrossdatasets, ineachofthefouragecategories.VGCanalysesforeachexaminationandacrosstheagecategoriesaresummarisedinfigures26to28.

(TPF-truepositivefraction;Lower-LowerVGCcurve;Upper-UpperVGCcurve)

Figure26:ChestVGCanalysisperagegroupA-<1year;B-1<5yearsold;C-5<10yearsold;D-10>16yearsold

As observed in Figure 26, the image quality of chest examination with the new



111

“post” exposure conditions obtained AUCVGC values higher than 0.5 in all agegroups.TheagegroupwiththehighestAUCVGCvalue(0.69)was<1year.


Figure27:AbdomenVGCanalysisperagegroupA-<1year;B-1<5yearsold;C-5<10yearsold;D-10>16yearsold

AsobservedinFigure27,theimagequalityofabdomenexaminationwiththenew“post”exposureconditionsobtainedAUCVGCvaluesequalorhigher than0.5 inallagegroups.TheagegroupwiththehighestAUCVGCvalue(0.62)was1-<5years.


Figure28:PelvisVGCanalysisperagegroupA-<1year;B-1<5yearsold;C-5<10yearsold;D-10>16yearsold

As observed in Figure 28, the image quality of pelvis examination with the new



112

“post”exposureconditionsobtainedAUCVGCvaluesequalorhigher than0.5 inallagegroups.TheagegroupwiththehighestAUCVGCvalue(0.62)was1-<5years.

Looking at the VGC analysis (considering all anatomical image criteria), all theradiographic images acquired by using the new exposure parameters for chest,abdomen and pelvis examinations obtained an AUCVGC value higher than 0.5 andtherefore, the four radiologists considered the imagequalityat leastequal to theexaminationsperformedwiththe“pre”exposureparameters.

ThesameVGCanalysiswasperformedforeachoftheanatomicalcriteriachosenbythe radiologists according to the European Guidelines on Quality Criteria forDiagnostic Radiographic Images in Paediatrics (table 25) for chest, abdomen andpelvisexaminationsandforeachagegroup.

Table25:VGCanalysisbyanatomicalcriterion

Imagequalityscoresforchestexaminationsscoredhigherinallfourimagecriteriaand in all age groups. Age group <1 showed the highest scores for the criteria“visually sharp reproduction of the diaphragm and costo-phrenic angles” and“reproduction of the spine and paraspinal structures and visualisation of theretrocardiaclungandthemediastinum”(AUCVGC=0.69).

Image quality scores for abdomen examinations scored higher in all four imagecriteria and in all age groups. Age group 1-<5 showed the highest scores for thecriteria “Visualisation of the psoas outline consistent with age and depending onbowelcontent”(AUCVGC=0.71).

Imagequalityscoresforpelvisexaminationsscoredhigherinallthreeimagecriteria



113

and in all age groups. Age group 1-<5 showed the highest scores for the criteria“Reproduction of the necks of the femora which should not be distorted byforeshorteningorexternalrotation”(AUCVGC=0.67).

Basedontheobjectiveandsubjective imageanalysis itcanbeconcludedthattheimagequalityobtainedwiththenewexposurefactorsgenerallyimproved.

Theresults fromboth theobjectiveandsubjective imageassessmentvalidate thenewexposurefactorsproposedtotheradiologydepartmentofHospitalC.

ThenewexposurefactorssignificantlyreducedbothESAKandKAPvalues,withoutaffectingimagequality(ingeneralimagequalityevenimproved).



114

§



115

3.5 Assessingtheuseofelectroniccroppinginplainimaging

Asexplained,propercollimationshouldbemadebeforetheexposureratherthanusing electronic cropping of the image after the exposure to avoid exposure ofextraneousbodyparts(Mooreetal.,2012).TheASRTWhitePaper(Herrmannetal.,2012) recommends that cropping should not be used as replacement for beamrestrictionthatcanbeachievedthroughphysicalcollimationoftheX-rayfieldsize.

It is consideredbad radiographypractice todigitally crop images insteadofusingproper collimation of the beam, as this leads to unnecessary radiation doseexposure to the patient. In addition, proper collimation of the examinations willalsocontributetonoisereductiononimages(Nyathietal.,2010).

Toevaluatetheuseofthepost-processingcroptoolindailypractice,theirradiatedandcroppedareas,availableinDICOMheadersonthePACSsystemfrom100chest,79 abdomen and 66 pelvis examinations were retrospectively analysed andcompared(table26).

Table26:irradiatedversuspostprocessedimagearea

examination age*group*(years)

mean*irradiated*

area*(cm2)

mean*post*processed*area*(cm2)

%*cropped*area

<1 260 210 19%

18<5 419 365 13%

58<10 650 575 12%

108<16 1038 921 11%

168<18 1296 1156 11%

<1 270 223 17%

18<5 592 500 15%

58<10 894 753 16%

108<16 1265 1067 16%

168<18 1385 1192 14%

<1 193 150 22%

18<5 314 237 24%

58<10 707 530 25%

108<16 1148 889 23%

168<18 1317 1030 22%

Chest

Abdomen

Pelvis



116

According to thedatacollected, the irradiatedarea is significantlyhigher (p<0.05,paired samples T-test) than the area needed to perform the analysed X-rayexaminations.

Theoverexposedareawasalsoverifiedinallpatientagegroups.Thisindicatesthatthecroppingtoolisbeingusedonaregularbasisandthereforecontributestotheincreaseofpatientdoseandscatteredradiation.

Forchestexaminations,theaveragepercentageofcroppedarearangedfrom11to19%.Thecroppingwashighestinpatientagegroup<1year.

Forabdomenexaminations, theaveragepercentageofcroppedarearanged from14to17%.Thecroppingwashighestinpatientagegroup<1year.

Forpelvisexaminations,theaveragepercentageofcroppedarearangedfrom22to25%,withthehighestcroppinginpatientagegroup5-<10years.

Results demonstrate that the use of post processing tools, such as electroniccollimation,hidesanunnecessaryoverexposurethatcouldbesubstantiallyreducedifabiggeremphasisweregiventoanappropriatecollimationbeforeexposure.

TheseresultswereshownanddiscussedwiththeradiographersandradiologistsoftheradiologydepartmentofHospitalC.Afterthediscussionswiththedepartmentteam, radiographers showed a strong commitment to use patient anatomicallandmarksmoreeffectivelytoensureproperandadequatebeamcollimation.



117

3.6 Limitationsofsection3

It is important to highlight that this study would benefit from using additionalanthropomorphicpaediatricphantoms(especiallyfornew-bornand10years).Thiswouldhavegiventhepossibilitytoprovidemoreobjectiveexposurecriteriaforallage groups. However, the model used, group discussion with radiologists andradiographers,combinedwithathoroughliteraturereview,allowedtodevelopnewexposureconditionsthatwereconsideredadequate.



118

§



119

4 ImpactoftheoptimisationprogrammeonpatientdosesFollowingthemodeldescribedinfigure22andafteroptimisationoftheexposurecriteria for chest, abdomen and pelvis plain radiography, a final round of patientdoseexposurecollectionwascarriedoutwiththeobjectivetomeasuretheimpactoftheoptimisationprogrammeonpatientdoses.

4.1 Material andmethods to assess the impact of optimisation onpatientdoses

InNovember2014,theradiologydepartmentofHospitalCstartedtousethenewexposure criteria for chest, abdomen and pelvis plain radiography. To determinethepostoptimisationDRLSs,datahasbeencollectedusingthesamemethodologyasdefined inphase1of thisstudyaswellasthesameequipment.Exposuredataandpatientweightwerecollectedfrom30patientsforeachagegroupandforeachexam(chest,abdomenandpelvis).

It is important to note that all the data collected was from properly referred,clinically justified exams and none of them was repeated due to inadequateexposure conditions and therefore all considered clinically valid by theradiographer,theradiologistandthemedicalreferrer.

Considering theuseof thenewexposurecriteria thathavebeenproposed to theradiologydepartmentforchest,abdomenandpelvisplainradiography,twomajorbenefitswereexpected:a)aharmonisationofpractice;b)asignificantreductionofExT,KAPandESAKvaluesbetweenphase1andthepostoptimisation(post)resultsforchest,abdomenandpelvis.

As for theharmonisationof practice, all radiographersofHospital C followed thenewexposurecriteriaandconsideredacommonguidancefortheproceduresverypositiveandbeneficial.



120

§



121

4.2 Resultsoftheimpactofoptimisationonpatientdoses

Figure29showsthecomparisonbetweenphase1andpostoptimisationExTmeanvaluesforchestplainradiography.

33,3#

7,9#

26,3#

9,9#

28#

6,1#

33,5#

7,4#

22#

6,7#

0#

5#

10#

15#

20#

25#

30#

35#

40#

phase#1# post# phase#1# post# phase#1# post# phase#1# post# phase#1# post#

<1# 15<5# 55<10# 105<16# 165≤18#

ExT#(ms)#

Figure29:ExposureTime(ms)valuesforchestplainradiography:phase1vspostoptimisation

Usingthepostoptimisationexposurecriterialedtoasignificantreduction(p<0.05,independent sample T-student test) of ExT in all age groups. The reduction was76%,62%,78%,78%,70%,respectivelyforagegroups<1,1-<5,5-<10,10-<16,16-≤18.

Figure30showsthecomparisonbetweenphase1andpostoptimisationExTmeanvaluesforabdomenplainradiography.

13,6%

2,7%

11%8,5%

18,2%

8,4%

46,2%

11,6%

0%

11,7%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

phase%1% post% phase%1% post% phase%1% post% phase%1% post% phase%1% post%

<1% 14<5% 54<10% 104<16% 164≤18%

ExT%(ms)%

Figure30:ExposureTime(ms)valuesforabdomenplainradiography:phase1vspostoptimisation



122

Usingthepostoptimisationexposurecriteria,ledtoasignificantreduction(p<0.05,independentsampleT-studenttest)ofExT.Thereductionwas80%,23%,54%,75%,respectivelyforagegroups<1,1-<5,5-<10,10-<16.

Figure31showsthecomparisonbetweenphase1andpostoptimisationExTmeanvaluesforpelvisplainradiography.

7,9$6$

9,3$ 8,4$

14$12$

23,8$

13,2$

49$

14,3$

0$

10$

20$

30$

40$

50$

60$

phase$1$ post$ phase$1$ post$ phase$1$ post$ phase$1$ post$ phase$1$ post$

<1$ 15<5$ 55<10$ 105<16$ 165≤18$

ExT$(ms)$

Figure31:ExposureTime(ms)valuesforpelvisplainradiography:phase1vspostoptimisation

Usingthepostoptimisationexposurecriterialedtoasignificantreduction(p<0.05,independentsampleT-studenttest)ofExT.Thereductionwas24%,10%,14%,45%,71%,respectivelyforagegroups<1,1-<5,5-<10,10-<16,16-≤18.



123

The KAP and ESAK values measured with the new proposed exposure criteriafollowed,asexpected,thesamereductiontendencyasobservedwiththeExT.

Tables27 to28 showtheKAPP75,ESAKP75andP75variationofphase1comparedwiththepostoptimisationvaluesforchest,abdomenandpelvisplainradiographyineachagegroup.

The KAPP75 and ESAKP75 values of the post optimisation phase are lower than inphase1forallthreeplainradiographyproceduresandinallagegroups,indicatingalowerpatientdoseexposure.

Table27:KAPP75,ESAKP75andP75variationvaluesforchestplainradiography:phase1vspostoptimisation(agegroups)

P75P75$

variation P75P75$

variation

phase)1 13 49

post 9 34

phase)1 22 52

post 10 40

phase)1 35 57

post 14 52

phase)1 68 73

post 41 60

phase)1 73 67

post 57 62

exposure)criteria

age)groups)(years)

!29%

!55%

!60%

16=≤18

<1

1=<5

5=<10

10=<16

ESAK)(µGy)KAP)(mGy.cm2)

!9%

!18%

!7%

!40%

!22%

!31%

!23%

Using thepostoptimisationexposure criteria for chestplain radiography reducedtheKAPP75valuesby22to60%.TheKAPP75reductionwashighest inagegroup5-<10.TheESAKP75 valueswere reducedby7 to31%,with thehighest reduction inagegroup<1(table27).



124

Table28:KAPP75,ESAKP75andP75variationvaluesforabdomenplainradiography:phase1vspostoptimisation(agegroups)

P75P75$

variation P75P75$

variation

phase)1 34 77

post 20 64

phase)1 72 110

post 47 77

phase)1 250 415

post 76 101

phase)1 1267 992

post 170 126

phase)1 n.a. n.a.

post 237 177

$35%

$70%

KAP)(mGy.cm2) ESAK)(µGy)

$16>≤18

<1

1><5

5><10

10><16 $87%

$

$17%

$30%

$76%

$87%

exposure)criteria

age)groups)(years)

$41%

Using the post optimisation exposure criteria for abdomen plain radiographyreduced theKAPP75valuesby35 to87%.TheKAPP75 reductionwashighest inagegroup 10-<16. The ESAKP75 values were reduced by 17 to 87%, with the highestreductioninagegroup10-<16(table28).

Table29:KAPP75,ESAKP75andP75variationvaluesforpelvisplainradiography:phase1vspostoptimisation(agegroups)

P75P75$

variation P75P75$

variation

phase)1 19 100

post 14 76

phase)1 30 113

post 28 99

phase)1 84 192

post 51 110

phase)1 267 379

post 55 120

phase)1 876 1204

post 93 164

exposure)criteria

age)groups)(years)

!26%

!7%

!39%

16=≤18

<1

1=<5

5=<10

10=<16


!43%

!68%

!86%

!79%

!89%

!24%

!12%



125

Using thepostoptimisationexposurecriteria forpelvisplain radiography reducedtheKAPP75valuesby7to89%.TheKAPP75reductionwashighest inagegroup16-≤18.TheESAKP75valueswerereducedby12to86%,withthehighestreductioninagegroup16-≤18(table29).

Tables30 to32 showtheKAPP75,ESAKP75andP75variationofphase1comparedwiththepostoptimisationvaluesforchest,abdomenandpelvisplainradiographyineachweightgroup.

The KAPP75 and ESAKP75 values of the post optimisation phase are lower than inphase1forallthethreeplainradiographyproceduresandinallweightgroups.

Table30:KAPP75,ESAKP75andP75variationvaluesforchestplainradiography:phase1vspost

optimisation(weightgroups)

P75P75$

variation P75P75$

variation

phase)1 21 140

post 9 26

phase)1 16 155

post 10 35

phase)1 28 199

post 15 46

phase)1 131 228

post 32 58

phase)1 145 189

post 57 67


!77%

!75%

!65%

!76%

!61%

!81%

!77%

≥50

<5

5A<15

15A<30

30A<50

exposure)criteria

weight)groups)(kg)

!57%

!38%

!46%

Using thepostoptimisationexposure criteria for chestplain radiography reducedtheKAPP75valuesby38to76%.TheKAPP75reductionwashighestinweightgroup30-<50.TheESAKP75valueswerereducedby65to81%,withthehighestreductioninweightgroup<5(table30).



126

Table31:KAPP75,ESAKP75andP75variationvaluesforabdomenplainradiography:phase1vspost


P75P75$

variation P75P75$

variation

phase)1 30 70

post 10 56

phase)1 35 82

post 20 65

phase)1 138 199

post 61 81

phase)1 248 285

post 203 113

phase)1 1445 1332

post 225 160

exposure)criteria

weight)groups)(kg)

!67%

≥50

<5

5@<15

15@<30

30@<50

!43%

!56%

KAP)(mGy.cm2) ESAK)(µGy)

!88%

!18%

!84%

!20%

!21%

!59%

!60%

Using the post optimisation exposure criteria for abdomen plain radiographyreducedtheKAPP75valuesby18to84%.TheKAPP75reductionwashighestinweightgroup ≥50. The ESAKP75 values were reduced by 20 to 88%, with the highestreductioninweightgroup≥50(table31).



127

Table32:KAPP75,ESAKP75andP75variationvaluesforpelvisplainradiography:phase1vspost


P75P75$

variation P75P75$

variation

phase)1 239 480

post 15 34

phase)1 37 125

post 18 85

phase)1 120 216

post 45 110

phase)1 314 327

post 75 152

phase)1 1164 1217

post 79 156


!49%

!54%

!87%

!76%

!93%

!93%

!32%

≥50

<5

5A<15

15A<30

30A<50

exposure)criteria

weight)groups)(kg)

!94%

!51%

!63%

Using thepostoptimisationexposurecriteria forpelvisplain radiography reducedtheKAPP75valuesby51to94%.TheKAPP75reductionwashighestinweightgroup<5.TheESAKP75 valueswere reducedby32 to93%,with thehighest reduction inweightgroup<5(table32).

Considering the post optimisation data analysis one can conclude that the twomajorbenefitsthatwereexpected:a)aharmonisationofpractice;b)asignificantreductionofExT,KAPandESAKvaluesbetweenphase1andthepostoptimisation(post)resultsforchest,abdomenandpelviswereachieved.Itisimportanttonotethat the quality of the images made with the new exposure criteria has beenevaluatedbypaediatricradiologistsandconsideredclinicallyacceptable.



128

§



129

5 PostoptimisationDRLsAftertheoptimisationprocesscarriedoutatHospitalC,newDRLvaluesbyageandweightgroupsweredefined.ItisimportanttohighlightthatsinceNovember2014all radiographersofHospitalChavebeenusingthenewexposurecriteriadefinedduringthedevelopmentofthisthesis.

5.1 NewDRLsbyagegroup

Table 33 shows the new optimised KAP and ESAK DRLs for chest PA/AP plainradiographyforeachagegroup,consideredastheP75value.

Table33:NewKAP&ESAKvaluesforchestAP/PA(byagegroups)


<1 9 3 15 8 34 20 40 30

10<5 10 5 18 9 40 25 50 35

50<10 14 10 24 13 52 26 60 44

100<16 41 20 56 35 60 27 68 51

160≤18 57 29 59 49 62 32 85 57

age7groups7(years)

KAP7(mGy.cm2) ESAK7(µGy)

Table 34 shows the new optimised KAP and ESAK DRLs for abdomen plainradiographyforeachagegroup,consideredastheP75value.

Table34:NewKAP&ESAKvaluesforabdomenAP(byagegroups)


<1 20 8 21 10 64 15 66 60

11<5 47 10 48 27 77 19 85 70

51<10 76 50 85 68 101 70 103 85

101<16 170 89 188 129 126 90 165 117

161≤18 237 198 265 217 177 98 185 139

age6groups6(years)




130

Table35showsthenewoptimisedKAPandESAKDRLsforpelvisplainradiographyforeachagegroup,consideredastheP75value.

Table35:NewKAP&ESAKvaluesforpelvisAP(byagegroups)


<1 14 8 15 10 76 22 80 75

1/<5 28 12 35 22 99 75 107 97

5/<10 51 28 58 50 110 90 112 105

10/<16 55 32 65 54 120 95 130 112

16/≤18 93 60 101 83 164 145 169 156

age7groups7(years)


ThesenewDRLswerepresentedtotheradiographersandradiologistsandwillbeusedbytheradiologydepartmentofHospitalCasfuturereferencefordoseauditprogrammes.



131

5.2 NewDRLsbyweightgroup

Table 36 shows the new optimised KAP and ESAK DRLs for chest PA/AP plainradiographyforeachweightgroup,consideredastheP75value.

Table36:NewKAP&ESAKvaluesforchestAP/PA(byweightgroups)


<5 9 4 11 7 26 20 30 22

5/<15 10 3 15 8 35 24 40 32

15/<30 15 8 24 11 46 26 50 40

30/<50 32 10 56 21 58 27 60 48

≥50 57 30 59 48 67 32 85 55

weight:groups:(kg)

KAP:(mGy.cm2) ESAK:(µGy)

Table 37 shows the new optimised KAP and ESAK DRLs for abdomen plainradiographyforeachweightgroup,consideredastheP75value.

Table37:NewKAP&ESAKvaluesforabdomenAP(byweightgroups)


<5 10 8 10 9 56 15 63 30

5/<15 20 8 21 18 65 19 85 60

15/<30 61 18 70 48 81 57 102 74

30/<50 203 60 265 92 113 82 185 102

≥50 225 125 245 193 160 98 182 132

weight8groups8(kg)




132

Table38showsthenewoptimisedKAPandESAKDRLsforpelvisplainradiographyforeachweightgroup,consideredastheP75value.

Table38:NewKAP&ESAKvaluesforpelvisAP(byweightgroups)


<5 15 8 15 12 34 22 35 28

5/<15 18 10 29 13 85 75 107 76

15/<30 45 18 50 28 110 81 110 99

30/<50 75 32 95 55 152 90 163 115

≥50 79 50 101 60 156 95 169 133

weight:groups:(kg)

KAP:(mGy.cm2) ESAK:(µGy)

ThesenewDRLswerepresentedtotheradiographersandradiologistsandwillbeusedbytheradiologydepartmentofHospitalCasfuturereferencefordoseauditprogrammes.



133

6 DiscussionSeveral research questions,major and specific objectiveswere formulated at thebeginningofthisthesis(page26).Adesignstructureforthestudywasdevelopedinordertoaccomplishthedesideratumconceivedforthisthesis(figure22,page95).Thissectiondiscussesthemainresultsachieved.

6.1 Aboutpatientcharacteristics

Inthelasttwoyears,anthropometricanddosedatafrom9,935paediatricpatientswas obtained in three dedicated paediatric hospitals of Portugal. To the bestknowledgeoftheauthors, thisstudyhasthe largestsampleofpaediatricpatientssubmittedtoaplainradiographyprocedurethatwasprospectivelyanalysed.

Thepatientsweremeasuredintermsofweight,heightandanatomicalthicknessofthe exposed structure, which allowed to propose values to define a “standardpatient”ineachagegroup,morefrequentlyusedinliterature(table6),asatooltoharmonisedatacollection for future studiesand todecreaseuncertaintiesdue tothe wide variation in patient characteristics, also identified by other authors(Gfirtner, Kaplanis, Moores, Schneider, & Vassileva, 2010): The lack ofharmonisation makes the attempts of comparing findings a very difficult, if notimpossibletask.

Median values of patient chest, abdomen and pelvis thickness, measured in oursample (table6,page77), are similar to thosepublished in literature (Hartetal.,2000;Wambanietal.,2013).Theresults fromthis thesiswill contributeto futurestudies, such as the development of a DRL curve as a function of the patientanatomical projection thickness and or weight, as proposed by some authors(Kiljunen et al., 2007) or for the development of new radiographic pre-setparameters,basedonpatient’s thickness,asproposedbyotherauthors (Zhangetal.,2013).

6.2 Aboutexposureparametersofphase1

Resultsof this thesishaveshowna largespread in theexposureparametersusedfor the same exam and age group, due to different choicesmade at the time ofexposure(tables7to9,pages78-80).Thesameresultswerefoundinotherstudies(Kristien Smans et al., 2008; Sonawane et al., 2011), illustrating a clear need forstandardisationinthisdomain.



134

The values of kV, mAs, ExT and SSD found in our results, combined with aninconsistent choice regarding the use of grid, AEC, protective shielding and X-raybeam limitation, are not according to International Guidelines (EuropeanCommission, 1996a; Herrmann et al., 2012; International Atomic Energy Agency,2012;InternationalCommissiononRadiologicalProtection,2013).

The authors corroborate the statement from Smans et al, that, if the exposuresettingsarenotusedaccordingtotheguidelines,therewillbealargeinfluenceonpatientdosimetry(KristienSmansetal.,2008).Thisinfluencewasidentifiedinourresults, showing an increase in patient doses due to inadequate use of exposuresettings.Partofthisproblemisrelatedtothefactthatpre-setsintheradiographicequipmentinstalledbythevendorarenotadaptedtopaediatricpatientsaswellasto a clear lack of education and training regarding theuseof digital systems andhowtomaximisetheirusetothebenefitofthepatient.

Regardingtheuseofprotectiveshielding,specificallytheprotectionofgonadswithlead, the recommendations given to the department took into consideration therecommendations from literature (Bardo et al., 2009; Fawcett & Barter, 2009;Herrmannetal.,2012).



135

6.3 AboutnationalDRLs

ThisisthefirstPortuguesestudythatprospectivelyanalysedpatientexposuredataandanthropometriccharacteristics ina largescaleandwithdatacollectedduringeachindividualprocedure.

Asalreadymentioned, theheterogeneityofmethodologies found in literature fordatacollectiontoestablishDRLsmakesthecomparisonofresultsaverychallengingtask. This challenge is even bigger when related to paediatric DRLs, due to thelimited number of studies available associated with the huge variety of patientcharacteristics.

Another important challenge is the potential confusion caused by the units andsubunits used in literature by different authors to report dose quantities. Forexample,todefineESAKvaluesauthorsmayuse:mGy,mGyorGy.TodefineKAPvalues,authorsmaypresentthevaluesinmGy.m2,dGy.cm2Gy.m2ormGy.cm2.ItiseasyforthereadertocompareESAKvalues,howeverforKAPitcanbedifficultandcreateconfusion,especiallywhenthedatacollectionismadethroughsurveys.

Tofacilitatetheworkofradiographersandradiologists,aconversiontableforKAPunits and subunits as the one presented in table 39 should be available in theradiologydepartments.

Table39:ConversionfactorsforKAPunits

1cGy.cm2=1µGy.m2

1µGy.m2=10mGy.cm2

1Gy.m2=10000000mGy.cm2

1dGy.cm2=100mGy.cm2

ItisimportanttonotethattheunitsdisplayedattheX-rayequipmentmonitorarenotalwaysthesameastheonessenttothePACS.Forexample,atHospitalsBandC,theequipmentconsoledisplaysKAPvaluesindGy.cm2,howeverthevaluesenttothePACSthroughtheDICOMheaderisinmGy.cm2,whichis100timessmallerthantheunitdisplayedon themonitor.This isobviously something tobecorrected toavoidaconfoundingelementforradiographers,radiologistsandmedicalphysicists.

Forpaediatricplain radiography theauthors recommend theuseof the followingsubunitswith the aim to harmonise the presentation of results in future studies:µGy forESAKandmGy.cm2 forKAP.Using thesesubunitswillavoid largedecimalnumbers of exposure quantities with several zeros (e.g. 0.0001) and will thusfacilitatecomparisonofresults.



136

The heterogeneity of patient grouping found in literature is also a considerablelimitationwhen itcomestodatacomparison.Also,moststudiesdonotallowthereader to identifyandcharacterise thewholespectrumofexposureconditionsasdescribedinsection6.2.

In the following tables the results of the DRLs obtained in this study will becomparedwithdifferentpublishedreferencevaluesforchest,abdomenandpelvisprojections.

Table40:ComparisonofvaluesforchestAP/PAplainradiographyESAKP75(µGy)withotherpublished

data

age$groups$(years)

present$study

Kostova$et$al$(2015)

Billinger$et$al$(2010)

EUR$16261$(1999)

Irish$DRL$(2004)

Hart$et$al$(2000)

Wambani$et$al$(2013)

Roch$et$al$(2013)

Vaño$et$al$(2008)

<1 145 30 52 80 57 50 50 80 51

1N<5 181 40 62 53 70 60 100 56

5N<10 209 70 82 100 66 120 70 200 91

10N<16 234 70 85 88 90 122

16N≤18 88

TheESAKP75valuesforchestplainradiographyobtainedinourstudyarehigherthanthe ones published in the literature and in the “European guidelines on qualitycriteria for diagnostic images in paediatrics” (Billinger et al., 2010; EuropeanCommission, 1996a; Hart et al., 2000; Kostova-Lefterova et al., 2015; MedicalCouncil Ireland, 2004;Roch&Aubert, 2013; EVañoet al., 2008;Wambani et al.,2013).



137

Table41:ComparisonofvaluesforabdomenplainradiographyESAKP75(µGy)withotherpublisheddata

age$groups$(years)

present$study


EUR$16261$(1999)

Irish$DRL$(2004)

Hart$et$al$(2000)


Vaño$et$al$(2008)

Roch$et$al$(2013)

<1 70 130 700 330 400 80 210

1K<5 191 387 752 500 130 401 1000

5K<10 198 1000 800 170 947 1500

10K<16 583 1200 200 2288

16K≤18 1258

RegardingtheESAKP75valuesforabdomenplainradiography,ourstudyshowsthelowestvalueswhencomparedwithotherpublishedresultsbesidesthestudyfromWambanietall(2013).

Table42:ComparisonofvaluesforpelvisplainradiographyESAKP75(µGy)withotherpublisheddata

age$groups$(years)

present$study

EUR$16261$(1999)

Irish$DRL$(2004)

Hart$et$al$(2000)


Vaño$et$al$(2008)

Roch$et$al$(2013)

<1 125 200 265 500 100 191 200

1J<5 158 475 600 120 673 900

5J<10 232 900 807 700 250 998 1500

10J<16 624 892 2000 360 2815

16J≤18 1204

As regards the ESAKP75 values for pelvis plain radiography, our study shows thelowestvalueswhencomparedwithotherpublishedresultsbesidesthestudyfromWambanietall(2013).

Table43:ComparisonofvaluesforchestplainradiographyKAPP75(mGy.cm2)withotherpublisheddata

age$groups$(years)

present$study

Smans$et$al$(2008)


Kiljunen$et$al$(2007)

Roch$et$al$(2013)

<1 13 88 23 22 30

1A<5 19 189 26 26 50

5A<10 60 233 37 28 70

10A<16 134 395 73 47

16A≤18 94

The KAPP75 values for chest plain radiography of our study are the lowest in agegroup <1 and 1-<5. In age groups 5-<10 and 10-<16 the KAPP75 values are



138

approximately 2 to 3 times higher than the study presented by Kiljunen et al(Kiljunenetal.,2007).

Table44:ComparisonofvaluesforabdomenplainradiographyKAPP75(mGy.cm2)withotherpublisheddata

age$groups$(years)

present$study

Smans$et$al$(2008)


Roch$et$al$(2013)

<1 25 21

1?<5 84 49 110 350

5?<10 140 102 700

10?<16 442 237

16?≤18 1401

TheKAPP75valuesforabdomenplainradiographyofourstudyarehigherinallagegroups when compared with the study showing the lowest values: Smans et al(KristienSmansetal.,2008).

Table45:ComparisonofvaluesforpelvisplainradiographyKAPP75(mGy.cm2)withotherpublisheddata

age$groups$(years)

present$study

Smans$et$al$(2008)

Roch$et$al$(2013)

<1 29 19 40

1><5 75 174 200

5><10 143 174 400

10><16 585 687

16>≤18 839

Looking at the KAPP75 values for pelvis plain radiography, our study shows thelowestvalueswhencomparedwithotherpublishedresults,exceptagegroup<1.

AlthoughDRL results are presented by patientweight group, it is not possible tocompare the results, since there are no studies published with patient weightgrouping,asrecommendedbythePiDRLproject(Damilakis,2015).

It is important to highlight that in terms of dosimetry, patientweight groups aremore appropriate (Kristien Smans et al., 2008), but in practice, the age of thepatient ismoreeasily obtainable. Toweighpaediatric patients at the timeof theradiologicalprocedureisnotonlytimeconsumingfortheprocedureworkflow,butrequires the radiology department to have different types of scales available, aspatient age ranges from new born to 18 years old. One option to surpass thisbottleneckwouldbetorecommendtheintegrationoftheweightdataatthetimeofreferral.



139

Asoutlinedabove,publishedresultsforpaediatricexaminationsarescarceandvarywidely as regards the methodology of collection and data presentation. Thisdemonstrates an urgent need to harmonise practices and to implementoptimisationprogrammes,especiallynowthatthegreatmajorityofhospitalshavedigital systems and therefore the possibility to decrease patient dose exposurewithout interfering with image quality. Professional societies and regulatoryauthoritiesshouldraiseawarenessonaregularbasisamongradiologydepartmentsandhealthprofessionalsabout the importanceofDRLdatacollection forabetterimplementationofdoseoptimisation.

6.4 Abouttheoptimisationtests

This study is, to theknowledgeof theauthors, the first tocombineobjectiveandsubjective analysis of radiological images and was based partly on a conceptualframework presented by some authors (Magnus Båth, Håkansson, Hansson, &Månsson,2005).

The objective analysis using anthropomorphic phantoms allowed comparing theimpactonimagequalityoftheexposurecriteriausedatthehospitalsiteforchest,abdomen and pelvis procedures, by using several images produced with varyingexposurecriteria.Withthephantomtestresultsfromsection3.2.2(page99)ithasbeen demonstrated that the new proposed exposure criteria did not affect theobjectiveimagequalityandallowedsignificantdosereductionineachexam.

To validate the new proposed exposure criteria, a subjective analysis of clinicalimageswasmadebyfourpaediatricradiologists,usingtheEuropeanGuidelinesonQuality Criteria for Diagnostic Radiographic Images in Paediatrics (EuropeanCommission, 1996a). Data was analysed using the VGC curves. Several authorsconsider thismethod as appropriate tomeasure image quality, because it offershigh validity, as it is based on assessment of clinically relevant structures(InternationalAtomicEnergyAgency,2014;Ludewigetal.,2010).

The results (section 3.4, page 109) demonstrated that the new recommendedexposure conditions, produced even better image quality with a significant dosereduction formost agebands anddiagnostic imagequalitywas confirmed.Otherauthorshaveshownthesameresults(L.Martinetal.,2013).

In addition, all staff involved, including radiographers and radiologists,welcomedthe opportunity to collaborate and were proactive in implementing the newoptimised protocols. The authors consider that it is necessary and of utmostimportancetoincludemedicalphysicists inthisprocess,notonlytofulfilthelegal



140

requirements, but because their knowledge and expertise improves theoptimisationprocess.

Duringtheoptimisationphaseofthisstudy,theuseofelectroniccroppinginplainimaginghasbeenanalysedandasignificantoverexposedareawasfound(from11to 25% - table 26, page 115) in all exams, resulting in unnecessary radiationexposure to the patient. Some authors identified the same problem (Soboleski,Theriault,Acker,Dagnone,&Manson,2006;Tschauneretal.,2015),whichcanonlybesurpassedby raisingawarenessamongst radiographersandby identifyingnewanatomicallandmarksforcollimation.



141

6.5 About the impact of the optimisation programme on patientdose

Regardingtheimpactoftheoptimisationprogrammepresentedinsection4.2(page121), the results of this study show a clear reduction of patient dose withoutaffecting the image quality for chest, abdomen and pelvis plain radiographyprocedures.

Totheknowledgeoftheauthorsthestudiesontheoptimisationprocessandtheirimpact found in literature are only related to paediatric chest plain radiography.Some authors report the same reduction impact on patient dose for chest plainradiography after implementation of an optimisation programme (Carlander,Hansson,Söderberg,Steneryd,&Båth,2010;Dabin,Struelens,&Vanhavere,2014;Frayre et al., 2012; L.Martin et al., 2013; Rizzi et al., 2014; Sanchez Jacob et al.,2009), suggesting that such optimisation programmes should be carried out on aregular basis, especially when new X-ray equipment or post-processing tools areinstalled.

Itisimportanttohighlightthatthisstudyaswellasthosecitedweredevelopedindedicated paediatric centres where it is expected that health professionals aremore attentive to the needs of the paediatric population. As shown by someauthors (Suliman & Elawed, 2013) it would be of major benefit if optimisationprogrammesforpaediatricpatientswerealsoimplementedingeneralhospitals.



142

§



143

Conclusions1. PortugueseDRLs for paediatric plain radiography (chest PA/AP, abdomen and

pelvis) were established and compared with European guidelines and otherstudiespublishedintheliterature.

2. TheDRLs(describedasKAPP75andESAKP75)wereestablishedbypatientageandweight groups, following international recommendations and were obtainedbeforeandafteranoptimisationprogramme.

3. The optimisation programme showed a significant patient dose reductionimpactinallagegroups(<1,1-<5,5-<10,10-<16,16-≤18).Dosewasreduced:

a. by22to60%andby7to31%,respectivelyforKAPP75andESAKP75forchestplainradiography(table27,page123);

b. by35to87%andby17to87%,respectivelyforKAPP75andESAKP75forabdomenplainradiography(table28,page124);

c. by7to89%andby12to86%,respectivelyforKAPP75andESAKP75forpelvisplainradiography(table29,page124);

4. The optimisation programme showed a significant an expressive patient dosereduction impact in all weight groups (<5kg; 5-<15kg; 15-<30kg; 30-<50kg;≥50kg).Dosewasreduced:

a. by38to76%andby65to81%,respectivelyforKAPP75andESAKP75forchestplainradiography(table30,page125);

b. by18to84%andby20to88%,respectivelyforKAPP75andESAKP75forabdomenplainradiography(table31,page126);

c. by51to94%andby32to93%,respectivelyforKAPP75andESAKP75forpelvisplainradiography(table32,page127);

5. Duetothesamplesizeofthisstudy(9,935)andthefactthatallpatientswereweighedandmeasured,itwaspossibletorecommendDRLsbyweightgroups.

6. Thisworkallowedproposingnewandharmonisedexposureparameters(tables21 to 23, pages 105-107) for chest, abdomen and pelvis plain radiography,facilitating dose reduction by up to 94%with image quality beingmaintainedafterasubjectiveimagequalityanalysisusingtheVisualGradingCharacteristicsmethod.

7. Several corrections were made regarding the way the dose reductiontechnological features, available on the X-ray equipment, were used.Improvementswereintroducedespeciallyaboutguidanceregardingtheuseofautomatic exposure control and the ionisation chambers, the antiscatter grid



144

and theuseof additional filtration, leading topatient dose reductionwithoutcompromisingimagequality;

8. After the implementation of the optimisation programme, the following localDRLswereestablishedforHospitalC:

a. CHEST:

age$groups$(years)

KAPP75$(mGy.cm2)

ESAKP75$(µGy)

<1 9 34

1><5 10 40

5><10 14 52

10><16 41 60

16>≤18 57 62

weight'groups'(kg)

KAPP75'(mGy.cm2)

ESAKP75'(µGy)

<5 9 26

5@<15 10 35

15@<30 15 46

30@<50 32 58

≥50 57 67

Anaveragereductionof41%and18%wasachievedforKAPP75andESAKP75respectively(agegroups)

Anaveragereductionof55%and75%wasachievedforKAPP75andESAKP75respectively(weightgroups)

b. ABDOMEN:

age$groups$(years)

KAPP75$(mGy.cm2)

ESAKP75$(µGy)

<1 20 64

1><5 47 77

5><10 76 101

10><16 170 126

16>≤18 237 177

weight'groups'(kg)

KAPP75'(mGy.cm2)

ESAKP75'(µGy)

<5 10 56

5A<15 20 65

15A<30 61 81

30A<50 203 113

≥50 225 160

Anaveragereductionof58%and53%wasachievedforKAPP75andESAKP75respectively(agegroups).

Anaveragereductionof54%and50%wasachievedforKAPP75andESAKP75respectively(weightgroups).



145

c. PELVIS:

age$groups$(years)

KAPP75$(mGy.cm2)

ESAKP75$(µGy)

<1 14 76

1=<5 28 99

5=<10 51 110

10=<16 55 120

16=≤18 93 164

weight'groups'(kg)

KAPP75'(mGy.cm2)

ESAKP75'(µGy)

<5 15 34

5A<15 18 85

15A<30 45 110

30A<50 75 152

≥50 79 156

Anaveragereductionof48%and47%wasachievedforKAPP75andESAKP75respectively(agegroups).

Anaveragereductionof75%and63%wasachievedforKAPP75andESAKP75respectively(weightgroups).

9. The optimisation programme presented in this thesis has led to an increasedawarenessofradiationprotectionamongsthealthprofessionals.Paediatricplainradiography doses have been reduced, however additional optimisation isnecessaryforotherprocedures,followingthesamemethodologycarriedoutbya multidisciplinary skilled and well-trained team, including a medical physicsexpert.GoodpracticeisrecommendedbyEuropeanguidelinesandisamatterrequiringfurtherresearch.



146

§



147

ReferencesAAPMTaskGroup116.(2009).AAPMTaskGroup116.Anexposureindicatorfor

digitalradiography.AmericanAssociationofPhysicistsinMedicine(Vol.36).

AltoComissariadodaSaúde.(2009).ComissãoNacionaldeSaúdedaCriançaedoAdolescente.2004-2008.Lisboa.

Amaral,D.,Matela,N.,Pereira,P.,&Palha,R.F.(2008).Evaluationofpaediatricexposureparametersanddosesin5-10yearsoldchildren.ComparisonbetweenDigitalandnon-digitalacquisitionsinfrontalThoraxprojection.RevistaLusófonadeCiênciasETecnologiasDaSaúde,(5),152–164.

AmericanAssociationofPhysicistsinMedicine.(2006).AAPMreporno.93.AcceptanceTestingandQualityControlofPhotostimulableStoragePhosphorImagingSystems.

AssembleiadaRepública.ConstituiçãodaRepúblicaPortuguesa.VIIrevisãoConstitucional(2005).Portugal.

Bardo,D.M.E.,Black,M.,Schenk,K.,&Zaritzky,M.F.(2009).Locationoftheovariesingirlsfromnewbornto18yearsofage:Reconsideringovarianshielding.PediatricRadiology,39(3),253–259.http://doi.org/10.1007/s00247-008-1094-4

Barros,P.,&Simões,J.(2007).HealthSystemsreview.HealthSystemsinTransition.(S.Allin&E.Mossialos,Eds.)(Vol.9).WHO.

Båth,M.,Håkansson,M.,Hansson,J.,&Månsson,L.G.(2005).Aconceptualoptimisationstrategyforradiographyinadigitalenvironment.RadiationProtectionDosimetry,114(1-3),230–235.http://doi.org/10.1093/rpd/nch567

Båth,M.,&Månsson,L.G.(2007).Visualgradingcharacteristics(VGC)analysis:anon-parametricrank-invariantstatisticalmethodforimagequalityevaluation.TheBritishJournalofRadiology,80(951),169–176.http://doi.org/10.1259/bjr/35012658

Billinger,J.,Nowotny,R.,&Homolka,P.(2010).DiagnosticreferencelevelsinpediatricradiologyinAustria.EuropeanRadiology,20(7),1572–9.http://doi.org/10.1007/s00330-009-1697-7

Breckow,J.(2006).Linear-no-thresholdisaradiation-protectionstandardratherthanamechanisticeffectmodel.RadiationandEnvironmentalBiophysics,44,257–260.http://doi.org/10.1007/s00411-006-0030-y

Brennan,P.C.,McEntee,M.,Evanoff,M.,Phillips,P.,O’Connor,W.T.,&Manning,D.J.(2007).AmbientLighting:EffectofIlluminationonSoft-CopyViewingofRadiographsoftheWrist.AmericanJournalofRoentgenology,188(2),W177–W180.http://doi.org/10.2214/AJR.05.2048

Brosi,P.,Stuessi,A.,Verdun,F.R.,Vock,P.,&Wolf,R.(2011).CopperfiltrationinpediatricdigitalX-rayimaging:Itsimpactonimagequalityanddose.RadiologicalPhysicsandTechnology,4(2),148–155.



148

http://doi.org/10.1007/s12194-011-0115-4

Busch,H.P.,&Faulkner,K.(2006).Imagequalityanddosemanagementindigitalradiography:Anewparadigmforoptimisation.RadiationProtectionDosimetry,117(1),143–147.http://doi.org/10.1093/rpd/nci728

Bushong.(2012).RadiologicsciencesforTechnologists:physics,biology,andprotection(10thed.).Houston:Elsivier.

Butler,M.L.,Rainford,L.,Last,J.,&Brennan,P.C.(2010).Areexposureindexvaluesconsistentinclinicalpractice?Amulti-manufacturerinvestigation.RadiationProtectionDosimetry,139(1-3),371–374.http://doi.org/10.1093/rpd/ncq094

Carlander,A.,Hansson,J.,Söderberg,J.,Steneryd,K.,&Båth,M.(2010).Theeffectofradiationdosereductiononclinicalimagequalityinchestradiographyofprematureneonatesusingadual-sidereadouttechniquecomputedradiographysystem.RadiationProtectionDosimetry,139(1-3),275–80.http://doi.org/10.1093/rpd/ncq072

Carroll,E.M.,&Brennan,P.C.(2003).RadiationdosesforbariumenemaandbariummealexaminationsinIreland:Potentialdiagnosticreferencelevels.BritishJournalofRadiology,76(906),393–397.http://doi.org/10.1259/bjr/13457134

Chotas,H.G.,Dobbins,J.T.,&Ravin,C.E.(1999).Principlesofdigitalradiographywithlarge-area,electronicallyreadabledetectors:areviewofthebasics.Radiology,210(3),595–599.http://doi.org/10.1148/radiology.210.3.r99mr15595

CommitteetoAssessHealthRisksfromExposuretoLowLevelsofIonizingRadiation.(2006).HealthRisksfromExposuretoLowLevelsofIonizingRadiation.BEIRVIIPhase2.NationalAcademyofSciences.Washington.

Cook,J.,Kyriou,J.,Pettet,A.,Fitzgerald,M.,Shah,K.,&Pablot,S.(2001).KeyfactorsintheoptimizationofpaediatricX-raypractice.BritishJournalofRadiology,74,1032–1040.

Cowen,A.,Davies,A.,&Kengyelics,S.(2007).Advancesincomputedradiographysystemsandtheirphysicalimagingcharacteristics.ClinicalRadiology.http://doi.org/10.1016/j.crad.2007.07.009

Dabin,J.,Struelens,L.,&Vanhavere,F.(2014).Radiationdosetoprematurenew-bornsinthebelgianneonatalintensivecareunits.RadiationProtectionDosimetry,158(1),28–35.http://doi.org/10.1093/rpd/nct184

Dackiewicz,D.,Bergsneider,C.,&Piraino,D.(2000).Impactofdigitalradiographyonclinicalworkflowandpatientsatisfaction.JournalofDigitalImaging :TheOfficialJournaloftheSocietyforComputerApplicationsinRadiology,13(2Suppl1),200–201.http://doi.org/10.1007/BF03167662

Damilakis,J.(2015).PiDRL-EuropeanCommissionTenderProjectondiagnosticreferencelevelsinpaediatricimaging.InsightsintoImaging,6(S1),1–158.http://doi.org/10.1007/s13244-015-0386-0

Don,S.,Macdougall,R.,Strauss,K.,Moore,Q.T.,Goske,M.J.,Cohen,M.,…



149

Whiting,B.R.(2013).Imagegentlycampaignbacktobasicsinitiative:tenstepstohelpmanageradiationdoseinpediatricdigitalradiography.AJR.AmericanJournalofRoentgenology,200(5),W431–6.http://doi.org/10.2214/AJR.12.9895

Don,S.,Whiting,B.R.,Rutz,L.J.,&Apgar,B.K.(2012).Newexposureindicatorsfordigitalradiographysimplifiedforradiologistsandtechnologists.AmericanJournalofRoentgenology,199(December),1337–1341.http://doi.org/10.2214/AJR.12.8678

Doyle,P.,Gentle,D.,&Martin,C.J.(2005).Optimisingautomaticexposurecontrolincomputedradiographyandtheimpactonpatientdose.RadiationProtectionDosimetry,114(1-3),236–9.http://doi.org/10.1093/rpd/nch548

Edmonds,K.D.(2009).Diagnosticreferencelevelsasaqualityassurancetool.TheRadiographer,56(3),32–37.

Eng,J.(2013).ROCAnalysis-Web-basedCalculatorforROCCurves.

ESR,EFRS,ESPR,EFOMP,&STUK.(2013).TenderfortheProjectENER/D3/91-2013.EuropeanGuidelinesonDiagnosticReferenceLevelsforPaediatricImaging.

EuropeanCommission.(1996a).EUR16261-EuropeanGuidelinesonQualityCriteriaforDiagnosticRadiographicImagesinPaediatrics.Luxembourg.

EuropeanCommission.(1996b).EUR16260.Europeanguidelinesonqualitycriteriafordiagnosticradiographicimages.Luxembourg.

EuropeanCommission.CouncilDirective97/43/EURATOM,onhealthprotectionofindividualsagainstthedangersofionizingradiationinrelationtomedicalexposure.,180OfficialJournaloftheEuropeanUnion22–27(1997).

EuropeanCommission.(1999).RPNo109.GuidanceonDiagnosticReferenceLevels(DRLs)forMedicalExposures(Vol.5).Luxembourg.

EuropeanCommission.RegulationNo1901/2006oftheEuropeanParliamentandoftheCouncilonmedicinalproductsforpaediatricuse.,OfficialJournaloftheEuropeanUnion(2006).

EuropeanCommission.(2007a).CommissiontakeslegalactionagainstPortugalregardingsafetystandardsforionizingradiation.RetrievedFebruary24,2015,fromhttp://europa.eu/rapid/press-release_IP-07-1527_en.htm

EuropeanCommission.CouncilDirective2007/47/EC.amendingCouncilDirective90/385/EEContheapproximationofthelawsoftheMemberStatesrelatingtoactiveimplantablemedicaldevices,CouncilDirective93/42/EECconcerningmedicaldevicesandDirective98/8/ECconcerni,OfficialJournaloftheEuropeanUnion(2007).

EuropeanCommission.RecommendationoftheEuropeanParliamentandoftheCouncilof23April2008ontheestablishmentoftheEuropeanQualificationsFrameworkforlifelonglearning,OfficialJournaloftheEuropeanUnion(2008).

EuropeanCommission.(2008b).RPN.o154.EuropeanGuidanceonEstimatingPopulationDosesfromMedicalX-RayProcedures.Luxembourg.

EuropeanCommission.(2009).RPNo159.EuropeanCommissionGuidelineson



150

ClinicalAuditforMedicalRadiologicalPractices(Vol.93).Luxembourg.

EuropeanCommission.ConsolidatedversionoftheTreatyEstablishingtheEuropeanAtomicEnergyCommunity(2012).OfficialJournaloftheEuropeanunion.

EuropeanCommission.CouncilDirective2013/59/EURATOM,layingdownbasicsafetystandardsforprotectionagainstthedangersarisingfromexposuretoionisingradiation,57OfficialJournaloftheEuropeanUnion1–73(2013).

EuropeanCommission.(2013b).Invitationtotenderno.ENER/D3/91-2013foraservicecontractregardingEuropeanGuidelinesonDiagnosticReferenceLevelsforPaediatricImaging.

EuropeanCommission.(2014a).RadiationProtectionno175:GuidelinesonRadiationProtectionEducationandTrainingofMedicalProfessionalsintheEuropeanUnion(1sted.).Luxembourg:PublicationOfficeoftheEuropeanUnion.

EuropeanCommission.(2014b).RPNo180,part2.DiagnosticReferenceLevelsinThirty-sixEuropeanCountries.Luxembourg.

EuropeanUnion.TreatyofLisbon.AmendingthetreatyonEuropeanUnionandthetreatyestablishingtheEuropeanCommunity(2007).

Fawcett,S.L.,&Barter,S.J.(2009).Theuseofgonadshieldinginpaediatrichipandpelvisradiographs.BritishJournalofRadiology,82(977),363–370.http://doi.org/10.1259/bjr/86609718

Frayre,aS.,Torres,P.,Gaona,E.,Rivera,T.,Franco,J.,&Molina,N.(2012).Radiationdosereductioninaneonatalintensivecareunitincomputedradiography.AppliedRadiationandIsotopes :IncludingData,InstrumentationandMethodsforUseinAgriculture,IndustryandMedicine,71Suppl,57–60.http://doi.org/10.1016/j.apradiso.2012.04.015

Gagliardi,R.(1996).AHistoryoftheRadiologicalSciences.RadiologyCentennial,Inc.

Gfirtner,H.,Kaplanis,P.a,Moores,B.M.,Schneider,P.,&Vassileva,J.(2010).AstudyinEuropeofpatientdosimetryindiagnosticradiology:protocoldevelopmentandfindings.RadiationProtectionDosimetry,139(1-3),380–7.http://doi.org/10.1093/rpd/ncq025

Gibson,D.J.,&Davidson,R.a.(2012).ExposureCreepinComputedRadiography.ALongitudinalStudy.AcademicRadiology,19(4),458–462.http://doi.org/10.1016/j.acra.2011.12.003

González,A.,&Darby,S.(2004).RiskofcancerfromdiagnosticX-rays:estimatesfortheUKand14othercountries.TheLancet,363,1908–1909.

Gray,J.E.,Archer,B.R.,Butler,P.F.,Hobbs,B.B.,Mettler,F.a,Pizzutiello,R.J.,…Yaffe,M.J.(2005).Referencevaluesfordiagnosticradiology:applicationandimpact.Radiology,235(2),354–8.http://doi.org/10.1148/radiol.2352020016

Hakansson,M.,Svensson,S.,Zachrisson,S.,Svalkvist,A.,Bath,M.,&Mansson,L.G.(2010).VIEWDEX:anefficientandeasy-to-usesoftwareforobserver



151

performancestudies.RadiationProtectionDosimetry,139(1-3),42–51.http://doi.org/10.1093/rpd/ncq057

Hart,D.,Wall,B.,Shrimpton,P.,Bungay,D.,&Dance,D.(2000).ReferenceDosesandPatientSizeinPaediatricRadiology.NRPB.

Haus,A.(1996).TheAAPM/RSNAPhysicsTutorialforResidents:MeasuresofScreen-FilmPerformance.Radiographics,16(5),1165–1181.

Herrmann,T.L.,Fauber,T.L.,Gill,J.,Hoffman,C.,Orth,D.K.,Peterson,P.a,…Odle,T.G.(2012).WhitePaper:Bestpracticesindigitalradiography.AmericanSocietyofRadiologicTechnologists.

Huda,W.(2005).Thecurrentconceptofspeedshouldnotbeusedtodescribedigitalimagingsystems.Radiology,234(2),345–346.http://doi.org/10.1148/radiol.2342040760

InternationalAtomicEnergyAgency.(2007).DosimetryinDiagnosticRadiology:aninternationalcodeofpractice.IAEAseries.Vienna:IAEA.

InternationalAtomicEnergyAgency.(2011).AvoidanceofUnnecessaryDosetoPatientsWhileTransitioningfromAnaloguetoDigitalRadiology(IAEA-TECDO).Vienna:IAEA.

InternationalAtomicEnergyAgency.(2012).RadiationProtectioninPaediatricRadiology.(IAEA,Ed.).Viena:IAEA.

InternationalAtomicEnergyAgency.(2013).DosimetryinDiagnosticRadiologyforPaediatricPatients.

InternationalAtomicEnergyAgency.(2014).DiagnosticRadiologyPhysics.(D.Dance,S.Christofides,A.Maidment,I.Mclean,&K.Ng,Eds.).Vienna.

InternationalCommissiononRadiationUnits&Measurements.(2005).PatientDosimetryforXRaysusedinMedicalImaging(Report74).JournaloftheICRU,5(2),iv–vi.http://doi.org/10.1093/jicru/ndi018

InternationalCommissiononRadiologicalProtection.(1977).ICRPPublication26.RecommendationsoftheICRP.AnnalsoftheICRP,1(3).

InternationalCommissiononRadiologicalProtection.(1996).ICRPPublication73.RadiologicalProtectionandSafetyinMedicine.AnnalsoftheICRP.

InternationalCommissiononRadiologicalProtection.(2001).Diagnosticreferencelevelsinmedicalimaging:reviewandadditionaladvice.AnnalsoftheICRP,31(3),33–52.

InternationalCommissiononRadiologicalProtection.(2004).Managingpatientdoseindigitalradiology.AreportoftheInternationalCommissiononRadiologicalProtection.AnnalsoftheICRP,34(1),1–73.http://doi.org/10.1016/j.icrp.2004.02.001

InternationalCommissiononRadiologicalProtection.(2005).ICRPPublication99.Low-doseExtrapolationofRadiation-relatedCancerRisk.AnnalsoftheICRP,35.

InternationalCommissiononRadiologicalProtection.(2007).ICRPPublication103.The2007RecommendationsoftheInternationalCommissiononRadiological



152

Protection.AnnalsoftheICRP.

InternationalCommissiononRadiologicalProtection.(2012).ICRPPublication119.CompendiumofDoseCoefficientsbasedonICRPPublication60.AnnalsoftheICRP,36.

InternationalCommissiononRadiologicalProtection.(2013).ICRPPublication121.Radiologicalprotectioninpaediatricdiagnosticandinterventionalradiology.AnnalsoftheICRP,42(2),1–63.http://doi.org/10.1016/j.icrp.2012.10.001

InternationalElectrotechnicalCommission.(2000).IEC60580.Geneva:IEC2000.

InternationalElectrotechnicalCommission.(2005a).IEC61267.Geneva:IEC2005.

InternationalElectrotechnicalCommission.(2005b).IEC61267ed.2.0B.MedicaldiagnosticX-rayequipment-Radiationconditionsforuseinthedeterminationofcharacteristics.Geneva:IEC.

InternationalElectrotechnicalCommission.(2008).IEC62494-1ed1.0.Medicalelectricalequipment-ExposureindexofdigitalX-rayimagingsystems-Part1:Definitionsandrequirementsforgeneralradiography.Geneva:IEC.

InternationalElectrotechnicalCommission.(2015).IEC62220-1-1ed1.0.Determinationofthedetectivequantumefficiency–Detectorsusedinradiographicimaging.Geneva:IEC.

James,J.J.,Davies,a.G.,Cowen,a.R.,&O’Connor,P.J.(2001).Developmentsindigitalradiography:Anequipmentupdate.EuropeanRadiology,11(12),2616–2626.http://doi.org/10.1007/s003300100828

Järvinen,H.,Seuri,R.,Kortesniemi,M.,Lajunen,A.,Hallinen,E.,Savikurki-Heikkilä,P.,…Tyrväinen,E.(2015).Indication-basednationaldiagnosticreferencelevelsforpaediatricCT:anewapproachwithproposedvalues.RadiationProtectionDosimetry,165(1-4),86–90.http://doi.org/10.1093/rpd/ncv044

Johnston,D.A.,&Brennan,P.C.(2000).ReferencedoselevelsforpatientsundergoingcommondiagnosticX-rayexaminationsinIrishhospitals.BritishJournalofRadiology,73(868),396–402.http://doi.org/10.1259/bjr.73.868.10844865

Kiljunen,T.,Ja,H.,&Savolainen,S.(2007).DiagnosticreferencelevelsforthoraxX-rayexaminationsofpaediatricpatients.TheBritishJournalofRadiology,80(June),452–459.http://doi.org/10.1259/bjr/60918774

Knight,S.P.(2014).ApaediatricX-rayexposurechart.JournalofMedicalRadiationSciences,191–201.http://doi.org/10.1002/jmrs.56

Korner,M.,Weber,C.,Wirth,S.,Pfeifer,K.,Reiser,M.,&Treitl,M.(2007).AdvancesinDigitalRadiography :Physical.Radiographics,27(3),675–686.

Kostova-Lefterova,D.,Taseva,D.,Hristova-Popova,J.,&Vassileva,J.(2015).Optimisationofpaediatricchestradiography.RadiationProtectionDosimetry,2,1–4.http://doi.org/10.1093/rpd/ncv119

Kotter,E.,&Langer,M.(2002).Digitalradiographywithlarge-areaflat-paneldetectors.EuropeanRadiology,12(10),2562–2570.http://doi.org/10.1007/s00330-002-1350-1



153

Lança,L.,&Silva,A.(2013).DigitalImagingSystemsforPlainRadiography(1sted.).NewYork:Springer.http://doi.org/10,10007/978-1-4614-5067-2_1

Lima,J.J.(2009).TécnicasdediagnósticocomraiosX(2nded.).Coimbra:ImprensadaUniversidadedeCoimbra.

Lin,E.C.(2010).Radiationriskfrommedicalimaging.MayoClinicProceedings.MayoClinic,85(December),1142–1146;quiz1146.http://doi.org/10.4065/mcp.2010.0260

Linet,M.S.,Kim,K.P.,&Rajaraman,P.(2009).Children’sexposuretodiagnosticmedicalradiationandcancerrisk:Epidemiologicanddosimetricconsiderations.InPediatricRadiology(Vol.39).http://doi.org/10.1007/s00247-008-1026-3

Ludewig,E.,Richter,A.,&Frame,M.(2010).Diagnosticimaging-Evaluatingimagequalityusingvisualgradingcharacteristic(VGC)analysis.VeterinaryResearchCommunications,34(5),473–479.http://doi.org/10.1007/s11259-010-9413-2

Martin,C.J.,LeHeron,J.,Borrás,C.,Sookpeng,S.,&Ramirez,G.(2013).Approachestoaspectsofoptimisationofprotectionindiagnosticradiologyinsixcontinents.JournalofRadiologicalProtection,33,711–734.http://doi.org/10.1088/0952-4746/33/4/711

Martin,L.,Ruddlesden,R.,Makepeace,C.,Robinson,L.,Mistry,T.,&Starritt,H.(2013).Paediatricx-rayradiationdosereductionandimagequalityanalysis.JournalofRadiologicalProtection,33(3),621–33.http://doi.org/10.1088/0952-4746/33/3/621

McCarty,M.,Waugh,R.,McCallum,H.,Montgomery,R.J.,&Aszkenasy,O.M.(2001).Pediatricpelvicimaging:Improvementingonadshieldplacementbymultidisciplinaryaudit.PediatricRadiology,31(9),646–649.http://doi.org/10.1007/s002470100515

McMullan,G.,Chen,S.,Henderson,R.,&Faruqi,a.R.(2009).Detectivequantumefficiencyofelectronareadetectorsinelectronmicroscopy.Ultramicroscopy,109(9),1126–1143.http://doi.org/10.1016/j.ultramic.2009.04.002

MedicalCouncilIreland.(2004).DiagnosticReferenceLevelsPositionPaper.

Moore,Q.T.,Don,S.,Goske,M.J.,Strauss,K.J.,Cohen,M.,Herrmann,T.,…Lehman,L.(2012).ImageGently:usingexposureindicatorstoimprovepediatricdigitalradiography.RadiologicTechnology(Vol.84).

Mothiram,U.,Brennan,P.C.,Lewis,S.J.,Moran,B.,&Robinson,J.(2014).Digitalradiographyexposureindices:Areview.JournalofMedicalRadiationSciences,61(2),112–118.http://doi.org/10.1002/jmrs.49

Neofotistou,V.,Tsapaki,V.,Kottou,S.,Schreiner-Karoussou,a,&Vano,E.(2005).Doesdigitalimagingdecreasepatientdose?Apilotstudyandreviewoftheliterature.RadiationProtectionDosimetry,117(1-3),204–210.http://doi.org/10.1093/rpd/nci718

Nickoloff,E.L.,Lu,Z.F.,Dutta,A.K.,&So,J.C.(2008).Radiationdosedescriptors:BERT,COD,DAP,andotherstrangecreatures.Radiographics,28(5),1439–1450.http://doi.org/10.1148/rg.285075748



154

Nyathi,T.,Chirwa,T.F.,&VanDerMerwe,D.G.(2010).AsurveyofdigitalradiographypracticeinfourSouthAfricanteachinghospitals:Anilluminativestudy.BiomedicalImagingandInterventionJournal,6(1).http://doi.org/10.2349/biij.6.1.e5

OECD.(2011).NuclearLegislationinOECDCountries.Portugal.

OECD.(2013).HealthataGlance:OECDIndicators.OECDPublishing.

OECD.(2015).Infantmortalityrates(indicator).

Park,H.-S.,Cho,H.-M.,Jung,J.,Lee,C.-L.,&Kim,H.-J.(2009).ComparisonoftheImageNoisePowerSpectraforComputedRadiography.JournaloftheKoreanPhysicalSociety,54(1),236.http://doi.org/10.3938/jkps.54.236

Paulo,G.,Rocha,M.,Lavandeira,P.,Costa,M.,&Marques,L.(2010).Exposuretooperatorsduringinterventionalcardiology.InInsightsintoImaging(p.230).Vienna:Springer.http://doi.org/10.1007/s13244-010-00

Paulo,G.,&Santos,J.(2012a).BOASPRÁTICASDEPROTECÇÃORADIOLÓGICANUMLABORATÓRIODEHEMODINÂMICA.InPrémioBoasPráticasemSaúde.Lisboa:APDH.

Paulo,G.,&Santos,J.(2012b).LocalDRLsforInterventionalCardiology.InAPIC2012.Lisboa:APIC.

Paulo,G.,Santos,J.,Moreira,A.,&Figueiredo,F.(2011).Transitionfromscreen-filmtocomputedradiographyinapaediatrichospital:themissinglinktowardsoptimisation.RadiationProtectionDosimetry,147(1-2),164–7.http://doi.org/10.1093/rpd/ncr355

Paulo,G.,Vaño,E.,&Rodrigues,A.(2015).Diagnosticreferencelevelsinplainradiographyforpaediatricimaging:APortuguesestudy.Radiography.http://doi.org/10.1016/j.radi.2015.07.002

Petoussi-Henss,N.,Zankl,M.,Drexler,G.,Panzer,W.,&Regulla,D.(1998).CalculationofbackscatterfactorsfordiagnosticradiologyusingMonteCarlomethods.PhysicsinMedicineandBiology,43(8),2237–2250.http://doi.org/10.1088/0031-9155/43/8/017

Rehani,M.M.(2015).Lookingintofuture:challengesinradiationprotectioninmedicine.RadiationProtectionDosimetry,1–4.http://doi.org/10.1093/rpd/ncv071

Rizzi,E.,Emanuelli,S.,Amerio,S.,Fagan,D.,Mastrogiacomo,F.,Gianino,P.,&Cesarani,F.(2014).OptimizationofExposureConditionsforComputedRadiologyExamsinNeonatalIntensiveCare.OpenJournalofRadiology,04(March),69–78.http://doi.org/10.4236/ojrad.2014.41009

Roch,P.,&Aubert,B.(2013).Frenchdiagnosticreferencelevelsindiagnosticradiology,computedtomographyandnuclearmedicine:2004-2008Review.RadiationProtectionDosimetry,154(1),52–75.http://doi.org/10.1093/rpd/ncs152

SanchezJacob,R.,Vano-Galvan,E.,Vaño,E.,GomezRuiz,N.,FernandezSoto,J.M.,MartinezBarrio,D.,&Prieto,C.(2009).Optimisingtheuseofcomputed



155

radiographyinpediatricchestimaging.JournalofDigitalImaging,22(2),104–113.http://doi.org/10.1007/s10278-007-9071-2

Santos,J.(2014).AninvestigationintotheoptimisationofpaediatricPortugueseComputedTomographypractice.UniversityCollegeofDublin.

Santos,J.,Foley,S.,Paulo,G.,McEntee,M.F.,&Rainford,L.(2014).TheestablishmentofcomputedtomographydiagnosticreferencelevelsinPortugal.RadiationProtectionDosimetry,158,307–317.http://doi.org/10.1093/rpd/nct226

Schaefer-Prokop,C.M.,DeBoo,D.W.,Uffmann,M.,&Prokop,M.(2009).DRandCR:Recentadvancesintechnology.EuropeanJournalofRadiology,72,194–201.http://doi.org/10.1016/j.ejrad.2009.05.055

Schaefer-Prokop,C.,Neitzel,U.,Venema,H.W.,Uffmann,M.,&Prokop,M.(2008).Digitalchestradiography:anupdateonmoderntechnology,dosecontainmentandcontrolofimagequality.EuropeanRadiology,18(9),1818–30.http://doi.org/10.1007/s00330-008-0948-3

Seibert,J.A.,&Morin,R.L.(2011).Thestandardizedexposureindexfordigitalradiography:Anopportunityforoptimizationofradiationdosetothepediatricpopulation.InPediatricRadiology(Vol.41,pp.573–581).http://doi.org/10.1007/s00247-010-1954-6

Smans,K.,Struelens,L.,Smet,M.,Bosmans,H.,&Vanhavere,F.(2010).Cufiltrationfordosereductioninneonatalchestimaging.RadiationProtectionDosimetry,139(1-3),281–6.http://doi.org/10.1093/rpd/ncq061

Smans,K.,Vaño,E.,Sanchez,R.,Schultz,F.W.,Zoetelief,J.,Kiljunen,T.,…Bosmans,H.(2008).ResultsofaEuropeansurveyonpatientdosesinpaediatricradiology.RadiationProtectionDosimetry,129(1-3),204–10.http://doi.org/10.1093/rpd/ncn031

Soboleski,D.,Theriault,C.,Acker,A.,Dagnone,V.,&Manson,D.(2006).Unnecessaryirradiationtonon-thoracicstructuresduringpediatricchestradiography.PediatricRadiology,36(1),22–25.http://doi.org/10.1007/s00247-005-0016-y

Sonawane,aU.,SunilKumar,J.V.K.,Singh,M.,&Pradhan,aS.(2011).SuggesteddiagnosticreferencelevelsforpaediatricX-rayexaminationsinIndia.RadiationProtectionDosimetry,147(3),423–8.http://doi.org/10.1093/rpd/ncq458

Sprawls,P.(2015).Thephysicalprincipalsofmedicalimaging.RetrievedMarch25,2015,fromwww.sprawls.org

Suliman,I.I.,&Elawed,S.O.(2013).RadiationDoseMeasurementsforOptimisationofChestX-RayExaminationsofChildreninGeneralRadiographyHospitals.RadiationProtectionDosimetry,156(3),310–314.http://doi.org/10.1093/rpd/nct073

Sun,Z.,Lin,C.,Tyan,Y.,&Ng,K.-H.(2012).Optimizationofchestradiographicimagingparameters:acomparisonofimagequalityandentranceskindosefordigitalchestradiographysystems.ClinicalImaging,36(4),279–86.http://doi.org/10.1016/j.clinimag.2011.09.006



156

Sun,Z.,Winder,J.R.,Kelly,B.E.,Ellis,P.K.,Kennedy,P.T.,&Hirst,D.G.(2004).AssessmentofVIEimagequalityusinghelicalCTangiography:invitrophantomstudy.ComputerizedMedicalImagingandGraphics,28(1-2),3–12.http://doi.org/10.1016/j.compmedimag.2003.09.001

Terrass,R.(1995).ThelifeofEdC.Jerman:ahistoricalperspective.RadiologicTechnology,66(5),291–298.

Tschauner,S.,Marterer,R.,Gübitz,M.,Kalmar,P.I.,Talakic,E.,Weissensteiner,S.,&Sorantin,E.(2015).EuropeanGuidelinesforAP/PAchestX-rays:routinelysatisfiableinapaediatricradiologydivision?EuropeanRadiology.http://doi.org/10.1007/s00330-015-3836-7

Tubiana,M.(2005).Dose-effectrelationshipandestimationofthecarcinogeniceffectsoflowdosesofionizingradiation:ThejointreportoftheAcadémiedesSciences(Paris)andoftheAcadémieNationaledeMédecine.InternationalJournalofRadiationOncologyBiologyPhysics,63(2),317–319.http://doi.org/10.1016/j.ijrobp.2005.06.013

Tubiana,M.,Feinendegen,L.E.,Yang,C.,&Kaminski,J.M.(2009).Thelinearno-thresholdrelationshipisinconsistentwithradiationbiologicandexperimentaldata.Radiology,251(1),13–22.http://doi.org/10.1148/radiol.2511080671

Ubeda,C.,Miranda,P.,&Vano,E.(2015).LocalpatientdosediagnosticreferencelevelsinpediatricinterventionalcardiologyinChileusingagebandsandpatientweightvalues.MedicalPhysics,42(2),615–622.http://doi.org/10.1118/1.4905116

Ubeda,C.,Vano,E.,Miranda,P.,&Leyton,F.(2012).PilotprogramonpatientdosimetryinpediatricinterventionalcardiologyinChile.MedicalPhysics,39(5),2424–30.http://doi.org/10.1118/1.3702590

UNSCEAR.(2013).UNSCEAR2013ReportVolumeII-Effectsofradiationexposureofchildren.NewYork.

USGovernment.FederalFood,Drug,andCosmeticAct,110thCongressPublicLaw85(2007).UnitedStatesofAmerica.

Vaño,E.,Fernadez,M.,Ten,J.I.,Prieto,C.,González,L.,Rodríguez,R.,&Heras,H.(2007).TransitionfromScreen-FilmtoDigitalRadiography:EvolutionofPatientRadiationDosesatProjectionRadiography.Radiology,243(2),461–466.

Vaño,E.,Martinez,D.,Fernandez,J.M.,Ordiales,J.M.,Prieto,C.,Floriano,A.,&Ten,J.I.(2008).Paediatricentrancedosesfromexposureindexincomputedradiography.PhysicsinMedicineandBiology,53(12),3365–3380.http://doi.org/10.1088/0031-9155/53/12/020

Wall,B.F.,Kendall,G.M.,Edwards,aa,Bouffler,S.,Muirhead,C.R.,&Meara,J.R.(2006).WhataretherisksfrommedicalX-raysandotherlowdoseradiation?TheBritishJournalofRadiology,79(940),285–94.http://doi.org/10.1259/bjr/55733882

Wambani,J.S.,Korir,G.K.,Korir,I.K.,&Kilaha,S.(2013).Establishmentoflocaldiagnosticreferencelevelsinpaediatricscreen-filmradiographyatachildren’shospital.RadiationProtectionDosimetry,154(4),465–76.



157

http://doi.org/10.1093/rpd/ncs270

Watson,D.J.,&Coakley,K.S.(2010).PaediatricCTreferencedosesbasedonweightandCTdosimetryphantomsize:localexperienceusinga64-sliceCTscanner.PediatricRadiology,40(5),693–703.http://doi.org/10.1007/s00247-009-1469-1

Willis,C.E.,&Slovis,T.L.(2005).TheALARAconceptinpediatricCRandDR:Dosereductioninpediatricradiographicexams-Awhitepaperconference.AmericanJournalofRoentgenology,184,373–374.http://doi.org/10.1007/s00247-004-1264-y

Zhang,M.,Liu,K.,Niu,X.,&Liu,X.(2013).Amethodtoderiveappropriateexposureparametersfromtargetexposureindexandpatientthicknessinpediatricdigitalradiography.PediatricRadiology,43(5),568–74.http://doi.org/10.1007/s00247-012-2555-3

Documents

OPTIMISATION AND ESTABLISHMENT OF DIAGNOSTIC ......To Professor Eliseo Vaño, Cathedratic Professor of Medical Physics of the Complutense University of Madrid, a world-renowned expert