质子治疗笔形束临床物理

ScanningBeams(ClinicalPhysics)MarkPankuchDirectorofMedicalPhysicsandDosimetryNorthwesternMedicineChicagoProtonCenterObjectives•IMPTOptimizationMethods•Robustness•OptimizationandEvaluation•AdditionalConcernswithIMPT•Motion•AdaptivePlanning•SummaryRadiationOncologyhasalwaysbeenaprogressiontobettertechnologyRadiationOncologyhasalwaysbeenaprogressiontobettertechnologyVMATStepandShootDynamicMLCProtonsareinthemiddleofthisevolutionApertureCompensatorIntensityModulatedProtonTherapy(IMPT)IMPT:Whatarewegaining?Collimation:LateralControlIntensityControlControlwithDepth3-DX-RayAperture/CompProtonIMRTIMPTSomeAdditionalBenefits•Decreasedsecondcancerrisks•InternalScatter•HeadLeakage:NeutronandphotonU.SchneiderandRHalg:FrontOncol2015IntensityModulationMethodsDistalEdgeTracking•SingleBraggPeaksdeliveredalongtheDistalEdgetargetonly(Dessy)•Severalincidentfieldsarenecessaryforauniformdose•RatioofenergydepositedintargettohealthilytissuesisthegreatestA.LomaxPhys.Med.Biol.44:(1999)IntensityModulationMethods2-DModulation•FixedRangeModulation•FieldSpecificRangeCompensatorA.LomaxPhys.Med.Biol.44:(1999)IntensityModulationMethods2.5-DModulation•VariableRangeModulation•SingleFieldOptimized(SFO)A.LomaxPhys.Med.Biol.44:(1999)IntensityModulationMethods3-DModulation•EachBraggpeakisoptimizedindividually•MultiFieldOptimized(MFO)A.LomaxPhys.Med.Biol.44:(1999)Spot/LayerPatternsforaSphereBeam’sEyeViewDoseDistributionHowcanweobtainthesecomplex,3-Dspotpatterns?Inverseplanningtechniques•Iterativeminimizationofanobjectivefunction.•Covertargetareas•Minimizedosetoorgansatrisk•Unlimitedpotentialforcomputingandmathematical“tricks”•Minimizingtheeffectsofpositionalerrorsandrangeerrorsdirectlyintothecostfunction•Includingeffectsofmotionintothecostfunction•MultiCriteriaOptimization(MCO)SingleFieldOptimized(SFO)/MultiFieldOptimized(MFO)MorecommonlyusedIMPTOptimizationMethodsSingleFieldOptimization(SFO)UniformDoseisdeliveredtotheentiretargetbyeachfieldindividuallyLesssparingofcriticalstructuresLesssensitivetoSet-up/RangeerrorsMultiFieldOptimization(MFO)Spotweightsofallfieldsareoptimizedtogether.Thespotweightofonefieldwillrelyonanotherfield’sdosetocreateanintegrateduniformtargetdoseBetterforsparingcriticalstructuresMoresensitivetoSet-up/RangeerrorsOAR100%ofDose100%ofDose+=(a)(b)(c)ClinicalRadiationOncology,3rdAdditionOAR100%ofDose100%ofDose+=(d)(e)(f)ClinicalRadiationOncology,3rdAddition+=SFOIntensityControl:MatchingFieldsIntensityModulationforFieldMatching20DoseProfilesatFieldJunction21PerpendicularExpansionAvoidageometricmissPhysicalDistance(cm)ParallelExpansionAvoidarangemissRadiobiologicalDepth(WaterEquiv.Thickness)MarginsforSetupandRangeUncertaintywithProtonsStaytunedforHakanNystrom:Uncertainties:3:30pmtodayOAR100%ofDose100%ofDoseOAR100%ofDose100%ofDoseThephysicististaskedtoconsiderQuantifytheeffectsof:•Non-idealset-up•Rangeuncertainty•Intra-fractionmotion•Respiratorymotion•Inter-fractionmotion•AnatomicalconsistencyRobustnessEvaluationQuantifythedifferencesinqualitybetweentheplannedandthedelivereddoseinthepresenceofuncertaintiesRobustPlanEvaluationincludes:•CalculationandEvaluationofmany“Worstcase”scenarios•SystematicoffsetofHUconversion(-3.5%,+3.5%)•Systematicoffsetofset-uperror(x=+/-5mm,y=+/-5mm,z=+/-5mm)Interpretationoftheevaluationresultsisdifficult•Inreallife,randomandsystematicerrorswillexist•Set-uperrorsarerandom•HUerrorsmaybesystematic,butnotsystematicallyoffsetoverthestudyset•ClearprioritizationoftargetcoverageandpotentialOARdosesmustbeunderstood.LomaxPhys.Med.Biol2008RobustnessEvaluationProspectiveRobustPlanningBeamAngleOptimizationpriortoSpotoptimization•Evaluatepath-lengthsandtheeffectofrangeerrorandset-upuncertainties•Conceptcanbeexpandedto4-DevaluationsCaoet.alMedPhys.2012Aug;39(8):5248–5256.MDAForSFO:BeamSpecificPTV:bs(PTV)P.ParketalIntJRadiatOncolBiolPhys.2012TheEffectofSet-upErrorsonSFOPlanParketal.IntJRadiatOncolBiolPhys.2012TheEffectofSet-upErrorsonSFOPlanParketal.IntJRadiatOncolBiolPhys.2012PlantoPTVPlantobsPTVAngularDependenceofBeam-SpecificPTVSetuperror:6mmInternalmotionerror:0mm(AVGCT)Rangeerror:3.5%oftotalWETSmearing:1cmbsPTVICTVMinimumCourtesyofLeiDongRobustOptimization•Addpenaltiesintothecostfunctionforrobustness•AllowtheplanningsystemtoscorerobustnessonaspottospotbasisANDhowonespotwilleffecttheoverallsensitivitytopotentialplandegradation.•Spotswith“poor”robustness(highsensitivitytoplandegradation)willbepenalizedbyiterativelydecreasing,andpotentially,eliminatingtheirintensity•ThereispotentialforBiologicalDoseOptimizationImportantNote:Robustoptimizationdoesnotalwaysguaranteearobustevaluation!RobustOptimization•RangeUncertaintiesandSet-upErrors1)Probabilisticapproach•Rangeandpositionalerrorsareparameterizedandincorporatedintotheobjectivefunction•spotrange/positionisarandomvariable2)Optimizeon“worstcase”scenariosUnkelbachPhys.Med.Bio.2007,MedPhys2009RobustOptimization(UnkelbachMed.Phys.2009)RobustOptimizationEvaluation(UnkelbachMed.Phys.2009)RobustOptimizationforOARRobustOptimizationforOARRobustOptimizationforOARThebalancingact