An evaluation of the effect of tube potential on clinical image ...

High tube potential, or 'high-kVp' techniques, which were established using traditional film-screen technology, are well-known and widely ... SkiptoArticleContent SkiptoArticleInformation JournalofMedicalRadiationSciencesVolume67,Issue4p.260-268 OriginalArticleOpenAccess Anevaluationoftheeffectoftubepotentialonclinicalimagequalityusingdirectdigitaldetectorsforpelvisandlumbarspineradiographs NicoleEPeacockBMedImag, CorrespondingAuthor NicoleEPeacockBMedImag [email protected] DepartmentofMedicalImaging,WesternHealth,Footscray,VIC,Australia SchoolofMedicine,FacultyofHealth,DeakinUniversity,WaurnPonds,VIC,Australia Correspondence NicoleEPeacock,DepartmentofMedicalImaging,WesternHealth,160GordonStreet,Footscray,VIC,Australia,3011.Tel:+610383456560;Fax:+610383456325;E-mail:Nicole.Peacock@wh.org.auSearchformorepapersbythisauthorAdamLStewardBAppSci, AdamLStewardBAppSci DepartmentofMedicalImaging,WesternHealth,Footscray,VIC,AustraliaSearchformorepapersbythisauthorPeterJRileyMSc, PeterJRileyMSc SchoolofMedicine,FacultyofHealth,DeakinUniversity,WaurnPonds,VIC,AustraliaSearchformorepapersbythisauthor NicoleEPeacockBMedImag, CorrespondingAuthor NicoleEPeacockBMedImag [email protected] DepartmentofMedicalImaging,WesternHealth,Footscray,VIC,Australia SchoolofMedicine,FacultyofHealth,DeakinUniversity,WaurnPonds,VIC,Australia Correspondence NicoleEPeacock,DepartmentofMedicalImaging,WesternHealth,160GordonStreet,Footscray,VIC,Australia,3011.Tel:+610383456560;Fax:+610383456325;E-mail:Nicole.Peacock@wh.org.auSearchformorepapersbythisauthorAdamLStewardBAppSci, AdamLStewardBAppSci DepartmentofMedicalImaging,WesternHealth,Footscray,VIC,AustraliaSearchformorepapersbythisauthorPeterJRileyMSc, PeterJRileyMSc SchoolofMedicine,FacultyofHealth,DeakinUniversity,WaurnPonds,VIC,AustraliaSearchformorepapersbythisauthor Firstpublished:03June2020 https://doi.org/10.1002/jmrs.403Citations:4AboutSectionsPDF ToolsRequestpermissionExportcitationAddtofavoritesTrackcitation ShareShare GiveaccessSharefulltextaccessSharefull-textaccessPleasereviewourTermsandConditionsofUseandcheckboxbelowtosharefull-textversionofarticle.IhavereadandaccepttheWileyOnlineLibraryTermsandConditionsofUseShareableLinkUsethelinkbelowtoshareafull-textversionofthisarticlewithyourfriendsandcolleagues.Learnmore.CopyURL SharealinkShareonFacebookTwitterLinkedInRedditWechat Abstract Introduction HighkVptechniques,15%or10-kVprules,arewell-knowndosereductionmethods.Traditionally,theuseofhightubepotential(i.e.increasedkVp)isassociatedwithdecreasedradiographiccontrastandoverallimagequality.RecentstudiessuggestcontrastandimagequalityarenotheavilyreliantonkVpwithdigitalsystems.Thisstudyaimstoassesstheeffectsofthehightubepotentialtechniqueonclinicalradiographicimagequalitywhenusingdigitalsystems,tovalidatehighkVpasadosesavingtechnique. Methods Aselectionofcomparablepelvisandlumbarspineradiographswerecollectedfromthehospital’spicturearchivingandcommunicationsystem(PACS),withtechnicalfactorsrecorded.Allclinicalradiographswereassessedby5seniorradiographersusinga15-pointvisualgradinganalysis(VGA)rubric. Results For40APpelvisradiographsand40laterallumbarspineradiographs,reductioninthedoseareaproduct(DAP)withhigherkVpisseen.AveragepelvisDAPat75 kVp = 14.06 mGy.cm2;85 kVp = 7.47 mGy.cm2.AveragelumbarspineDAPat80 kVp = 15.76 mGy.cm2;90 kVp = 14.83 mGy.cm2.ImagequalityandcontrastscoresshowednostatisticallysignificantdifferencebetweenthehighandlowkVpgroups(z = 0.06and0.12,respectively).AveragepelvisVGAscoreat75 kVp = 11.26;85 kVp = 12.55.AveragelumbarspineVGAscoreat80 kVp = 9.23;90 kVp = 10.64. Conclusions Thehightubepotentialtechniquesallowedforreducedpatientradiationdoseswhilstshowingnodegradationofdiagnosticimagequalityinaclinicalsetting.ThisstudysuccessfullyvalidatesthehighkVptechniqueasausefultoolforreducingpatientradiationdoseswhilstmaintaininghighdiagnosticimagequalityfordigitalpelvisandlumbarspineradiography. Introduction Diagnosticradiologyexaminationsinvolvingionisingradiationcarryanelementofriskandthepotentialtocauseharm.Inordertominimisetheserisksandensurepatientsafety,radiationexposuresusedduringradiologicalexaminationsshouldbekept‘aslowasreasonablyachievable’(ALARA).1Exposurefactorssetbyclinicalradiographersatthex-rayconsoleforeachexaminationdirectlyimpacttheresultantpatientradiationdose;therefore,optimisationofexposurefactorselectionisparamount.Withtheintroductionofdirectdigitalimagingsystems,radiographersarenolongerlimitedbytheneedtoachieveafixedfilm-screenopticaldensityandcanthereforenowmanipulateexposurefactorstoachieveimagequalityandpatientdoseoptimisation.Thereareanumberofoptimisationtechniquesdescribedintheliterature,includingbothhighandlowkVptechniques.Hightubepotential,or‘high-kVp’techniques,whichwereestablishedusingtraditionalfilm-screentechnology,arewell-knownandwidelydocumentedmethodsofreducingpatientradiationdoses.2-7ThehighkVpconceptfordosereductionhassincebeencarriedovertodigitalimaging.The‘15%’and‘10-kVp’rulesareusedtogovernthesetechniques,statingthatanincreaseintubepotential,measuredaskilovoltagepeak(kVp),by15%or10-kVprequiresthemilliamperageseconds(mAs)value(whichindicatestheproductofthetubecurrentandexposuretime)tobehalvedinordertokeepthedetectordoseconstant.2-4,8-10TherearealargenumberofstudiesintheliteraturethatconfirmhighkVp,lowmAstechniquesareeffectiveinreducingpatientradiationdoseswithdigitalimagingsystems.8,9,11-17Thesedosesavingabilities,however,arenotinfinite;somestudiesshowthatasthekVpreacheshighervalues(i.e.above100 kVp),thedosereductionsdecreaseandeventuallytaperoff.Intheir2014study,Reisetal.13showedadosereductionof18%whenthekVpwasincreasedfrom80to90 kVpbutonlya2%dosereductionwhenthekVpwasincreasedfrom100to110 kVp.ThissuggeststhereisalimittohowfarthehighkVpphenomenoncanextend. Traditionally,whenfilm-screensystemswereused,anincreaseinkVpmeantalossofimagecontrastandoverallimagequality.1,4,5,7,8,16Currently,thereisdisagreementwithintheliteratureabouthowkVpselectioneffectsimagequalityandcontrastondigitalimagingsystems.Somesources,includingBushong’stextbook‘RadiologicScienceforTechnologists:Physics,BiologyandProtection’,2statethatkVphasnoeffectonresultingimagecontrast.2,6,18,19Otherstudiesdisagree,withsomefindingreducedcontrastandimagequality1,5-7,9,20andothers,improvedimagequalitywithhigherkVpvalues.4LowkVptechniqueshavealsobeendescribed,asamethodofimagequalityoptimisationthroughexploitingthewidedynamicrangeofdigitalflat-paneldetectors.Aspreviouslydescribed,withmoderndirectdigitalimagingsystems,radiographersarenolongerlimitedbytheneedtoachieveafixedopticaldensityandthereforeareabletoincreasethedetectordoseforimprovedsignaltonoiseratios,whilstkeepingpatientradiationdosesthesame,byreducingkVpvalues.3,4,8Withthesedifferingdose/imagequalityoptimisationtechniquesandconflictinginformationseenincurrentpublishedliterature,questionsareraisedaboutthevalidityofsuchtechniquesandthebestmethodsforoptimisation.Asclinicalradiographersattempttobalancethetrade-offbetweenradiographicimagequalityandpatientradiationdoses,furtherinvestigationintotheseoptimisationtechniquesisrequiredtodeterminewhichtechniqueismostbeneficialinwhichspecificclinicalsituation. AimsandObjectives Theaimofthisretrospectiveauditstudywastoinvestigatetheeffectivenessofthehigh-kVptechniqueinaclinicalsettingwithdirectdigitalradiographicsystemswithflat-paneldetectors,forthereductionofpatientradiationdoseswhilstmaintainingdiagnosticimagequalityforpelvicandlumbarspineimaging.Theobjectiveswereto: QuantifiablyassesstheeffectsofincreasingkVpvaluesonradiographicimagequalitywhenthedetectordoseremainsconstant AssessthedosesavingcapabilitiesofhighkVptechniquesintermsofpatientradiationdoses Usereal,clinicalpelvicandlumbarspineradiographsforevaluationtoensureresultsarerelevantandapplicabletoclinicalradiographers. PreviousresearchhasshownpelvisimagingtohavesignificantpatientdosesavingresultswhenusinghighkVptechniques.6APpelvisprojectionswereincludedinthisstudytoensuredosesavingsassociatedwithusinghigherkVpvaluesdonotcausesignificantdegradationtotheresultingimagequality.Laterallumbarspineprojectionswerealsoincludedasarelativelythickbodyregion,pronetoreducedimagequalityfromscatterradiation,toensureahigherkVptechniquewouldnotresultinundiagnosticimagequality. Hypothesis WithincreasedkVpvaluesandfixeddetectorexposures,theauthorspostulatedthat: Radiographicimagequalitywouldnotbesignificantlycompromised Allimageswouldremaindiagnosticallyacceptable Patientradiationdoseswouldbereduced. Methods EthicalapprovalwassoughtandgrantedforthisretrospectivestudybyWesternHealth’sLowRiskEthicsPanelinJuly2018. Imagingsystems AllradiographsincludedinthisstudywereacquiredatWesternHealthusingthedepartment’sGEOptimaXR656DirectDigitalRadiographicUnitswith‘Flash-Pad’Flat-PanelWirelessDigitalDetectors,asinglepanelamorphoussilicondetectorwithaCesiumIodidescintillator(GEHealthcare,Chicago,USA). Datacollection ThestudywasdesignedasaretrospectiveanalysisofexistingradiologicalimagedataonWesternHealth’sFujifilmSynapsepicturearchivingandcommunicationsystem(PACS)(FujifilmMedicalSystems,Stamford,USA).Antero-posterior(AP)pelvisradiographsandlaterallumbarspineradiographswereanalysedovertimeperiodsof3and6 months,respectively(fromOctober2017toMarch2018).Atotalof217APpelvisimagesand91laterallumbarspineimageswereavailableonWesternHealth’sPACSwithinthesetimeframes.TechnicalinformationthatwasstoredonPACSwasrecordedforeachradiographthatmettheinclusioncriteria,includingkilovoltagepeak(kVp),milliamperage/sec(mAs),deviationindex(DI)anddoseareaproduct(DAP). Inclusioncriteria Asaretrospectivestudy,theauthorsfollowedastrictinclusioncriteriatoensurecomparableradiographsandpatienttype(i.e.thickness)werebeinganalysed.TheassumptionwasmadethatpatientthicknessshouldberelativelyconsistentacrossradiographsgeneratedusingthesamekVpiftherewasnomorethan25%variationinthemAsandnogreaterthan1.0differenceinthedeviationindex(DI).DIisaGEmeasurementofexposure,representingthevariationbetweentheactualexposureindex(EI)andthetargetEI,withthebase10logarithmicscale.EIisameasureoftheradiationreachingthedetectorand,whendividedby100,iscomparabletothedetectorentrancedosemeasurementintheunitµGy(whencalibrationconditionsareused).ThetargetEIvariesdependingonbodyregion,projectionandpatientsizeselected.TheEI,andsubsequentlytheDI,iscalculatedthroughacomplicatedcalculationworkflowthattakesintoaccountanatomy(throughananatomicregionidentificationalgorithm),kVp/mAs,filtration,grid,receptor,speed,detectorsensitivityandmedianimagecount;afurtherdescriptionofwhichisbeyondthescopeofthispaper.ADIof0isoptimal,withtheoptimalrangeofDIvaluesbetween−3and+2.25 ThefirstinclusioncriterionwasthatthekVpusedtoacquiretheradiographhadtomatcheitheraloworhighkVpvalue.Thesespecificvaluesweresetafterreviewingtrendsinexposureselectionoverthegiventimeperiodsofdatacollection.APpelvisimagesneededtobeacquiredat75or85kVpandlaterallumbarspineimagesat80or90kVp.ThemAsvaluesusedforacquisitionsneededtofallwithinsetparameters,withthemAsofthehigherkVpgroupdeliberatelysettohalfthatofthelowerkVpgroup,inkeepingwiththe‘10-kVp’ruledescribedpreviously.ForAPpelvisradiographsacquiredat75kVp,20mAs ∓ 25%wassetastheinclusionrange.ForAPpelvisradiographsacquiredat85kVp,10mAs ∓ 25%wassetastheinclusionrange.Forlaterallumbarspineradiographsacquiredat80kVp,80mAs ∓ 25%wassetastheinclusionrange.Forlaterallumbarspineradiographsacquiredat90kVp,40mAs ∓ 25%wassetastheinclusionrange.ThelastinclusioncriterionwasfortheradiographtohaveaDIvaluefallingbetween−0.5and + 0.5.Theseassumptionsonexposureselectionandpatientsizearebasedonclinicalassumptionsmadeattheconsolebyradiographersandareinkeepingwiththeclinicallybasedfocusofthisstudy.Theydohoweverpresentalimitationofthestudy,whichisdescribedindetailinthediscussion. Exclusioncriteria Nopaediatricimageswereincludedinthisstudy.NoradiographstakenonWesternHealth’sGEmobileX-raymachineswereincludedinthisstudy.Asthiswasaclinicalstudyusingreal,diagnosticradiographs,theauthorsmadesomeallowanceforoverlyingartefactsorpathologythatmayobscurelandmarksusedforimagequalityscoring,excludingasmallnumberofradiographs,examplesofwhichareshowninFigure 1. Figure1OpeninfigureviewerPowerPoint Examplesofexcludedradiographs(A)Grossoverlyingbowelgas(B)Severescoliosis.Anexclusioncriteriawereusedtoexcluderadiographssuchasthesefromthisstudy,withoverlyingartefactorpathologyobscuringvisualisationoflandmarksusedforVGAscoringtopreventthesenegativelyinfluencingVGAscores. Afterapplyingtheinclusionandexclusioncriteria,therewere26APpelvisradiographsacquiredat75 kVp,30APpelvisradiographsacquiredat85 kVp,22laterallumbarspineradiographsacquiredat80kVpand31laterallumbarspineradiographsacquiredat90 kVpeligibleforinclusion.Thesenumberswereroundedto20radiographspergroup,forevencomparisonsandeaseofimageanalysis. Imageanalysis AllimageswereviewedonaDellP3320LCDmonitor(Dell,RoundRock,USA)with1680x1050-pixelresolutionthroughtheFujifilmSynapsePACS(FujifilmMedicalSystems,Stamford,USA).Thesamemonitorwasusedbyeachreviewertolimitanyvarianceinthedisplay.Radiographswereindependentlyassessedby5seniorradiographers,eachwithatleast15 yearsofclinicalexperience.Assessorswereblindedtotheexposurefactorsusedforeachimage,whichwerepresentedinarandomisedorder.WindowingonPACSwaspermitted,tomimicconditionsofclinicalimageanalysisandcritique.Eachimagewasawardedascorefrom0to15,basedonavisualgradinganalysis(VGA)rubric,showninFigure 2,thatwasderivedfromtheCommissionofEuropeanCommunitiesGuidelinesforImageQualityCriteriaforDiagnosticRadiographicImages.26Theseguidelineshavebeenwidelyusedandadaptedforimagequalityassessmentsintheliterature,withanumberofthepreviouslymentionedhigh-kVpstudiesadoptingtheseguidelinesinsomecapacity.8,11,12,14-16 Figure2OpeninfigureviewerPowerPoint ExampleVGARubricfor(A)APpelvisprojection(B)Laterallumbarspineprojection.DerivedfromthecommissionofEuropeancommunitiesguidelinesforimagequalitycriteriafordiagnosticradiographicimages,26thevisualgradinganalysisrubricwasusedtoquantitativelyassessthecontrastandimagequalityofradiographsinthisstudy. Dataanalysis Thenon-parametricMann–WhitneyU-testwasusedtodeterminewhetherastatisticallysignificantdifferencebetweenVGAscoresexistedbetweenthetwogroups,highandlowkVp.Thistestallowedforastandardz-scoretobederivedwithacorrespondingPprobabilityvalue,throughrankingthescoresawardedandcalculatingtheranksumofscoresineachgroup.Az-scoregreaterthan1.96showsstatisticalsignificanceatthe95%confidenceinterval.27 IndividualDAPvalueswereusedasanestimateofmeanpatientradiationdosesforeachprojection,ineachindividualkVpgroup. Results VGAresults RawVGAdataforthe40pelvisand40lumbarspineradiographsareavailableinSupplementaryInformation.Theaveragescoreforeachindividualradiograph,forbothhighandlowkVpgroups,ispresentedinFigure 3.HighscoringpelvisandlumbarspineimagesareshowninFigures 4and5,respectively,withaverageVGAscoresforreference.ThemeanVGAscoreforAPpelvisat75kVpwas11.26andat85kVpwas12.55.ThemeanVGAscoreforlaterallumbarspinesat80kVpwas9.23andat90kVpwas10.64.TheMann–WhitneyU-testwasappliedtotheAPpelvisdata,lowvshighkVpgroups,givingaprobabilityvalueofP = 0.06.Thelaterallumbarspinedata,lowvshighkVpgroups,gaveaprobabilityvalueofP = 0.12.TheAPpelvisandlaterallumbarspinedatawerecombinedtoassesslowvshighkVpoverall,givingaz-scoreofz = 3.1andprobabilityvalueofP = 0.03. Figure3OpeninfigureviewerPowerPoint AverageVGAscoresof(A)APpelvisimages(B)Laterallumbarspineimages.ThemeanVGAscoreforeachindividualradiographisshowninthesegraphs.Whilstthereissomevarianceseen,moreoftenthannotthehigherkVpradiographs(purplelines)showhigherVGAscoresandthereforebetterimagequality,thanthelowerkVpgroup. Figure4OpeninfigureviewerPowerPoint APpelvisradiographs(A)LowkVpgroup#19(B)HighkVpgroup#11.ExamplesofhighVGAscoringAPpelvisradiographsfromthelowandhighkVpgroups. Figure5OpeninfigureviewerPowerPoint Laterallumbarspineradiographs(A)LowkVpgroup#3(B)HighkVpgroup#12.ExamplesofhighVGAscoringlaterallumbarspineradiographsfromthelowandhighkVpgroups DAPresults RawDAPdataforthe40pelvisand40lumbarspineradiographsareavailableinSupplementaryInformationandaregraphedforvisualrepresentationinFigure 6.ThemeanDAPvaluesascalculatedfortheAPpelviswere14.06 mGy.cm2at75 kVpand7.47 mGy.cm2at85 kVp.ThemeanDAPvaluesascalculatedforthelaterallumbarspinewere15.76 mGy.cm2at80 kVpand14.83mGy.cm2at90 kVp. Figure6OpeninfigureviewerPowerPoint DAPvaluesof(A)APpelvisimages(B)Laterallumbarspineimages.TheDAPvaluerepresentstheradiationdosedeliveredtothepatientineachprojection.ReducedradiationdosesareseeninthehigherkVpgroups(purplelines),withmorepronounceddosesavingsseeninthepelvicradiographs Discussion VGAresults TheprobabilityvaluesofP = 0.06andP = 0.12,fortheAPpelvisandlaterallumbarspineVGAresults,respectively,provethatthereisnostatisticallysignificantdifferencebetweentheranksumsoftheVGAscoresfromthehighandlowkVpgroupsforeitherprojection.TheseresultswereinkeepingwiththehypothesisthatincreasingkVpwouldnotsignificantlyreduceradiographicimagequality,andagreewithanumberofpublishedmaterials,thatwhendigitalsystemsareusedtheimagequalityandcontrastarenotheavilyreliantonthekVpvalue.6,18,19Interestingly,themeanVGAscoresforthehighandlowkVpgroupsshowedincreasedaveragescoresforthehigherkVpgroupsinbothprojections,byover1VGApoint.Whenalldatafrombothprojectionswerecombined,az-scoreof3.1showsastatisticallysignificantdifferencebetweenthehighandlowkVpgroups,favouringhighkVp. TheseresultssupportthefindingsofdeVries,whosecontrastdetailphantomimagingalsosawslightlyhigherimagequalityscoreswhenhigherkVpvalueswereused.4Althoughtheseresultsdisagreewithfindingsfromanumberofotherstudies,9,14,20-24theclinicalnatureofthisstudydesignandthespecificprojectionsusedmayhelpexplainthedifferences.BothFauber9andAndria20imagedphantoms(apelvisphantomandaLeedsTestObjectphantom,respectively)intheirresearchinstrictlaboratorysettings,meaningallconditionsofimageacquisitionwerekeptconstantandidenticaltissuethicknessandtissuetypewasusedforallacquisitions.Thisretrospective,clinicalstudydesigndidnotallowforcontroloverimageacquisitionconditions,norwasdatacollectedonpatientsizeorthickness.Whilsteveryeffortwasmadetoensurecomparableradiographswereincluded,thepotentialvariancefromanumberofinfluencingfactorscouldexplainthediscrepanciesinresults.WhilstGuo’sstudywasalsoaclinicalstudy,theirresearchlookedintothechangestoimagequalityofpaediatricchestradiographswithvaryingkVpvalues.14ChestradiographydeliberatelyutiliseshighkVpvalues,generallybetween110and125 kVp,toproducealowcontrastradiographallowingforoptimalvisualisationofthefinevascularmarkingsofthelungs.18Theprojectionsusedinthecurrentstudyarespecificallylookingatbonyanatomyandthereforehavedifferentgoalstochestimaging(i.e.highcontrastrequiredtovisualisebonytrabecularpatterns). Theresultsfromthecurrentstudyalsodisagreewithstudiesinvestigatinglow-kVp,variabledetectordosetechniques,thatfoundimprovedimagequalitywhenlowerkVpvalueswereused.Geijer5foundimprovedimagequalityandloweredeffectivepatientdoseswhenlumbarspinekVpvalueswerereducedfrom77to60 kVpfortheAPprojectionandfrom90to77 kVpforthelateral.Kuwahara24assessedimagequalityandlunglesionvisibilityonphantomchestradiographsvaryingfrom90to140 kVp.Imagesacquiredat90 kVpweredeemedtohavesuperiorimagequality,withnosignificantdifferenceseenbetweenlesionvisibility.Thesetwolow-kVpstudies,however,utilisedthemoremodernconceptofvariabledetectorexposureandalsoadjustedotherradiographicfactorstoachievetheseresults,withGeijer5increasingthesystemspeedfrom400to800andKuwahara24addingcopperfiltration. DAPresults Overall,forbothprojectionsstudied,resultssawdecreasedDAPvaluesandthereforedecreasedpatientradiationdoses,whenthehigherkVpvalueswereused.ThisagreeswithandvalidatesthehighkVptechniqueasadosesavingtoolondirectdigitalradiographicsystems.8,9,11-17Therewere,however,differencesintheamountofdosereductionobservedbetweeneachprojectionandregionofinterest.ForAPpelvisprojections,increasingthekVpfrom75to85sawa46.9%reductioninthemeanDAPvalue.Forlaterallumbarspineprojections,increasingthekVpfrom80to90sawonlya5.9%reductioninthemeanDAPvalue.Thisdifferencemaybeexplainedbydosesavingtaperoff,asdescribedbyReis;13askVpvaluesincreasethepercentageofdosesavingsdetectedbetweeneachsteplessens. Studylimitations Thisstudysolelyfocusesonthetraditionalfilm/screenconceptofa‘10-kVp’ruleforconstantdetectordoseandreducedpatientdoses.Thistechniqueisjustoneofmanyimagequality/doseoptimisationmethodsavailableinmoderndigitalradiography.Asdiscussedpreviously,theclinicalandretrospectivenatureofthisstudydesignlefttheauthorswithnocontroloverspecificimageacquisitionconditions,orinformationonpatientsizeorthickness.Thestudythereforereliedonradiographicparameterstoensuretheradiographsused,andpatientthicknessforeach,weresomewhatcomparable.TheDIvaluewasusedasonemeasureofstandardisation;however,manyvariablescaneffectitscalculation,suchascollimationsize,centringpointandpatientthickness.GiventhattheDIisinfluencedbycollimation,itshouldbenotedthatinthecaseoftheAPpelvisradiographs,automatedcollimationoccurstothe41 × 41 cmdetectorsizeandsothecollimationisassumedtobethesameforeachimage.Thiswasnotthecaseforthelaterallumbarspineimageswhichwerecollimatedwithinthedimensionsofthedetectorandthereforepresentasalimitationofthestudy. Althoughsomeallowancewasmadeforgrossartefactsorpathology,theclinicalradiographsusedmayhavehadsomeartefactorpathology(e.g.overlyingbowelgas)thatmadevisualisationofrelevantbonylandmarksdifficult,negativelyinfluencingtheVGAscores. Next,nosophisticatedradiationdosemeasurementsweretakenatthetimeofimageacquisition;dosemeasurementsweretakenfromtheestimatedDAPvaluerecordedontheGEsystem,whichwillhaveamarginforerror.DAPrecordingsfromthesystemcanbeinfluencedbythecollimatedarea,andthiscouldleadtoerroneousreadings.Asmentioned,thecollimatedfieldsizefortheAPpelvisimageswasconsistentbutthisisnotthecaseforthelaterallumbarspineimages. Lastly,themonitorusedforimagereviewingwasnotadiagnosticstandarddisplayusedforradiologyreporting.ThiscouldalsohavethepotentialtonegativelyinfluencetheVGAscores. Futuredirections Toconfirmresultsseeninthisretrospectivestudy,furtherresearchwithlargersamplesizesandmoreobserverstoscoreimageswouldbebeneficial.Differentregionsofinterestandradiographicprojections,especiallythosethatrequiresofttissuedetail,suchasabdominalx-rays,wouldalsobeusefultoinvestigate.AfterthepromisingVGAresultsofthisstudy,thepossibilityoffurtherincreasetokVpvalues,topushthelimitsofthisdosesavingphenomenon,couldalsobeconsidered.Withmanyimagequality/doseoptimisationtechniquesavailableindigitalradiography,furtherresearchcomparingfixedandvariabledetectorexposuremethodsandtheirabilitytoimprovediagnosticimagequalitywhilstdecreasingpatientradiationdosesisneeded. Conclusion ThisstudyallowedforquantifiableassessmentoftheeffectofincreasingkVpvaluesonradiographicimagequalityandassessmentofpatientradiationdosesforx-rayimagingofthelumbarspineandpelvisinaclinicalsettingusingdirectdigitaldetectors.TheresultssupportedthehypothesisthatincreasingkVpvalueswillnotsignificantlyreduceimagequality.Theimagequalityofallradiographswasdiagnosticallyacceptable,andoverall,areductioninpatientradiationdose(intermsofDAPvalues)wasobservedwhenhigherkVpvalueswereused.ThisstudysuccessfullyvalidatesthehighkVptechniqueasausefultoolforreducingpatientradiationdoseswhilstmaintaininghighdiagnosticimagequalityforpelvisandlumbarspineimaging. Acknowledgements TheauthorswishtothankClairedeBooy,ParveenSingh,JacquiAtallah,MagdalenaRybka,ShezardArnoldaandAssocProf.GiovanniMandaranofortheircontributions. ConflictofInterest Theauthorsdeclarenoconflictofinterest. SupportingInformation Filename Description jmrs403-sup-0001-Supinfo.docxWorddocument,13.2KB SupplementaryInformation.RawVGAresultsfor(a)APpelvisimages(b)Laterallumbarspineimages.Eachradiographwasscoredbyfiveindividualassessorsfrom0to15usingtheVGArubricforradiographiccontrastandimagequality. Pleasenote:Thepublisherisnotresponsibleforthecontentorfunctionalityofanysupportinginformationsuppliedbytheauthors.Anyqueries(otherthanmissingcontent)shouldbedirectedtothecorrespondingauthorforthearticle. References 1The2007RecommendationsoftheInternationalCommissiononRadiologicalProtection.AnnalsoftheICRP2007;37(2–4):9–34. GoogleScholar 2BushongS.RadiologicScienceforTechnologists:Physics,BiologyandProtection,10thedn.ElsevierMosby,Missouri,2013. GoogleScholar 3ChingW,RobinsonJ,McEnteeM.Patient-basedradiographicexposurefactorselection:asystematicreview.JMedRadiatSci2014;61(3):176–90. WileyOnlineLibraryPubMedGoogleScholar 4deVriesG,HoekB,PlankeelE,BijwaardH,HaarlemN.editors.Optimizingimagequalityatconstanteffectivedosebyadjustingtubevoltageforlumbarspineradiographyusingaflatpanelsystem.EuropeanCongressofRadiology,Vienna,2016. GoogleScholar 5GeijerH,NorrmanE,PerslidenJ.Optimizingthetubepotentialforlumbarspineradiographywithaflat-paneldigitaldetector.BrJRadiol2009;82(973):62–8. CrossrefCASPubMedWebofScience®GoogleScholar 6StewardA,TarolloM,MandaranoG.Conventionalfilm/screenprinciplesofexposureselectioninthedigitalage.RadiologicTechnologist2018;89(5):488–94. PubMedWebofScience®GoogleScholar 7ShanahanMC.ApilotstudyinvestigatingtwodosereductiontechniquesforAPlumbarspineradiographyusingdirectdosimetryandProjectionVR.Radiography(Lond)2017;23(3):222–8. CrossrefCASPubMedGoogleScholar 8RamanaiduS,StaMariaR,NgK,GeorgeJ,KumarG.Evaluationofradiationdoseandimagequalityfollowingchangestotubepotential(kVp)inconventionalpaediatricchestradiography.BiomedImagingIntervJ2006;2(3):e35. CrossrefCASPubMedGoogleScholar 9FauberT,CohenT,DempseyM.Highkilovoltagedigitalexposuretechniqueandpatientdosimetry.RadiolTechnol2011;82(6):501–10. PubMedGoogleScholar 10AllenE,HoggP,MaWK,SzczepuraK.Factorfiction:Ananalysisofthe10kVp‘rule’incomputedradiography.Radiography2013;19(3):223–7. CrossrefGoogleScholar 11CharnleyC,EnglandA,MartinA,TaylorS,BensonN,JonesL.Anoptionforoptimisingtheradiographictechniqueforhorizontalbeamlateral(HBL)hipradiographywhenusingdigitalX-rayequipment.Radiography2016;22(2):e137–e142. CrossrefWebofScience®GoogleScholar 12LorussoJR,FitzgeorgeL,LorussoD,LorussoE.Examiningpractitioners’assessmentsofperceivedaestheticanddiagnosticqualityandhighkVp-LowmAspelvis,chest,skull,andhandphantomradiographs.JMedImagingRadiatSci2015;46(2):162–73. CrossrefPubMedWebofScience®GoogleScholar 13ReisC,GonçalvesJ,KlompmakerC,etal.ImagequalityanddoseanalysisforaPAchestX-ray:ComparisonbetweenAECmodeacquisitionandmanualmodeusingthe10kVp‘rule’.Radiography2014;20(4):339–45. CrossrefGoogleScholar 14GuoH,LiuWY,HeXY,ZhouXS,ZengQL,LiBY.Optimizingimagingqualityandradiationdosebytheage-dependentsettingoftubevoltageinpediatricchestdigitalradiography.KoreanJRadiol2013;14(1):126–31. CrossrefPubMedWebofScience®GoogleScholar 15DohertyP,O’LearyD,BrennanPC.DoCECguidelinesunder-utilisethefullpotentialofincreasingkVpasadose-reducingtool?EurRadiol2003;13(8):1992–9. CrossrefCASPubMedWebofScience®GoogleScholar 16AxelssonB,PetersenU,WiltzHJ.Reductionofabsorbeddosebyadaptionofexposurefactorsandimageprocessing.ActaRadiol2001;42:592–8. WileyOnlineLibraryCASPubMedWebofScience®GoogleScholar 17BarbaJ,CulpM.CopperFiltrationandkVp:EffectonEntranceSkinExposure.RadiolTechnol2015;46(2):162–73. GoogleScholar 18BontragerK,LampignanoJ.TextbookofRadiographicPositioningandRelatedAnatomy,8thedn.ElsevierMosby,Missouri,2014. GoogleScholar 19ElhussienN.EffectofKVpinDigitalImageQuality.SudanUniversityofScienceandTechnology,Sudan,2015. GoogleScholar 20AndriaG,AttivissimoF,GuglielmiG,LanzollaA,MaioranaA,MangiantiniM.Towardspatientdoseoptimizationindigitalradiography.Measurement2016;79:331–8. CrossrefWebofScience®GoogleScholar 21UffmannM,NeitzelU,ProkopM,etal.Flat-panel–detectorchestradiography:Effectoftubevoltageonimagequality.Radiology2005;235(2):642–650. CrossrefPubMedWebofScience®GoogleScholar 22JonesA,AnsellC,JerromC,HoneyI.Optimizationofimagequalityandpatientdoseinradiographsofpaediatricextremitiesusingdirectdigitalradiography.TheBritishJournalofRadiology2015;99(1050):20140660. CrossrefWebofScience®GoogleScholar 23AngeliniL,PoglianiM,SantucciG,etal.X-rayacquisitionofextremities:Imagequalityanddoseoptimization.PhysicaMed2018;56(2):228–229. GoogleScholar 24KuwaharaC,AokiT,OdaN,etal.Optimalbeamqualityforchestflatpaneldetectorsystem:Realisticphantomstudy.EurRadiol2019;29(9):4538–4543. CrossrefPubMedWebofScience®GoogleScholar 25HealthcareGE.DetectorExposureIndicatorinGEX-RAYSystems.GeneralElectricCompany,Buckinghamshire,2014. GoogleScholar 26CarmichaelH,MacciaC,MooresB,OestmannJ,SchibillaH,TeunenD.Europeanguidelinesonqualitycriteriafordiagnosticradiographicimages.OfficeforOfficialPublicationsoftheEuropeanCommunities,Luxembourg,1996. GoogleScholar 27Mathematicav.11.2.0.0,WolframResearchInc(1997-2019). GoogleScholar CitingLiterature Volume67,Issue4December2020Pages260-268 FiguresReferencesRelatedInformation CloseFigureViewerReturntoFigurePreviousFigureNextFigureCaption DownloadPDF Thefulltextofthisarticlehostedatiucr.orgisunavailableduetotechnicaldifficulties. LogintoWileyOnlineLibrary EmailorCustomerID Password Forgotpassword? NEWUSER> INSTITUTIONALLOGIN> ChangePassword OldPassword NewPassword TooShort Weak Medium Strong VeryStrong TooLong PasswordChangedSuccessfully Yourpasswordhasbeenchanged Createanewaccount EmailorCustomerID Returninguser Forgotyourpassword? Enteryouremailaddressbelow. EmailorCustomerID Pleasecheckyouremailforinstructionsonresettingyourpassword. Ifyoudonotreceiveanemailwithin10minutes,youremailaddressmaynotberegistered, andyoumayneedtocreateanewWileyOnlineLibraryaccount. RequestUsername Can'tsignin?Forgotyourusername? Enteryouremailaddressbelowandwewillsendyouyourusername EmailorCustomerID Close Iftheaddressmatchesanexistingaccountyouwillreceiveanemailwithinstructionstoretrieveyourusername