MRITTTstar區(qū)別專題知識

MRIPhysics2:ContrastsandProtocolsChrisRorden,PaulMorganTypesofcontrast:ProtocolsStatic:T1,T2,PDEndogenous:T2*BOLD(‘fMRI’),DWExogenous:GadoliniumPerfusionMotion:ASLwww.fmrib.ox.ac.uk/~karla/1MRContrast–adefinitionWeusedifferentMRIprotocolsthataredominatedbydifferentcontrasts.Contrastsinfluencethebrightnessofavoxel.Forexample,water(CSF)isrelativelydarkinaT1-weightedscan,butrelativelybrightinaT2scan.2MRContrastFourtypesofMRcontrasts:StaticContrast:Sensitivetorelaxationpropertiesofthespins(T1,T2)EndogenousContrast:Contrastthatdependsonintrinsicpropertyoftissue(e.g.fMRIBOLD)Exogenouscontrast:Contrastthatrequiresaforeignsubstance(e.g.Gadolinium)Motioncontrast:Sensitivetomovementofspinsthroughspace(e.g.perfusion).3AnatomyofanMRIscanPlaceobjectinstrongstaticmagneticfield,then.TransmitRadiofrequencypulse:atomsabsorbenergyWaitListentoRadioFrequencyemissionduetorelaxationWait,Goto1Timebetweenset1and3isourEchoTime(TE)Timebetweenstep1beingrepeatedisourRepetitionTime(TR).TRandTEinfluenceimagecontrast.TimeTRTE4T1andT2definitionsT1-Relaxation:RecoveryRecoveryoflongitudinalorientation.‘T1time’referstointervalwhere63%oflongitudinalmagnetizationisrecovered.T2-Relaxation:DephasingLossoftransversemagnetization.‘T2time’referstointervalwhereonly37%oforiginaltransversemagnetizationispresent.50.2Contrast:T1andT2EffectsT1effectsmeasurerecoveryoflongitudinalmagnetization.T2referstodecayoftransversemagnetization.T1andT2varyfordifferenttissues.Forexample,fathasverydifferentT1/T2thanCSF.Thisdifferencecausesthesetissuetohavedifferentimagecontrast.T1isprimarilyinfluencedbyTR,T2byTE.CSF:LongT1Fat:ShortT1MagnetizationTR(s)0301CSF:LongT2Fat:ShortT2SignalTE(s)0016T1Effects:getthemwhiletheirdownConsiderveryshortTR:Fathasrapidrecovery,eachRFpulsewillgeneratestrongsignal.Waterhasslowrecovery,littlenetmagnetizationtotip.T1effectsexplainwhywediscardthefirstfewfMRIscans:thesignalhasnotsaturated,sothesescansshowmoreT1thansubsequentimages.Beforefirstpulse:

1Hinalltissue

stronglymagnetized.CSFFatAfterseveralrapidpulses:CSFhaslittlenetmagnetization,sothesetissuewillnotgeneratemuchsignal.7SignalDecayAnalogyAfterRFtransmission,wecandetectRFemissionEmissionatLarmorfrequency.Emissionsamplitudedecaysovertime.Analogoustotuningfork:frequencyconstant,amplitudedecays8RelaxationAfterRFabsorptionends,protonsbegintoreleaseenergyEmissionatLarmorfrequency.Emissionsamplitudedecaysovertime.Differenttissuesshowdifferentratesofdecay.‘FreeInductionDecay’(FID).Strongestsignalimmediatelyaftertransmission.MostsignalwithshortTE.WhynotalwaysuseshortTE?9TEandT2contrastSignalsfromalltissuedecayswithtime.Signaldecaysfasterinsometissuesthanothers.Optimalcontrastbetweentissuewhentheyemitrelativelydifferentsignals.GrayMatter:SlowDecayWhiteMatter:FastDecaySignalTE(s)0.2Contrast:differencebetweenGMandWMsignalSignalTE(s)0.2OptimalGM/WMcontrast10OptimalcontrastOptimalTEwilldependonwhichtissuesyouwishtocontrastGraymatter

vsWhitematterCSF

vsGraymatterSignalTE(s)0.211T2:DephasingRFpulsesetsphase.Initially,everythinginphase:maximumsignal.Signalsgraduallydephase=signalisreduced.Sometissueshowsmorerapiddephasingthanothertissue.TimeCSFFat…12T1andT2contrastsEveryscanisinfluencedbybothT1andT2.However,byadjustingTEandTRwecandeterminewhicheffectdominates:T1-weightedimagesuseshortTEandshortTR.Fatbright(fastrecovery),waterdark(slowrecovery)T2-weightedimagesuselongTEandlongTR:theyaredominatedbytheT2Fatdark(rapiddephasing),waterbright(slowdephasing).ProtondensityimagesuseshortTEandlongTR:reflecthydrogenconcentration.AmixtureofT1andT213T2vsT2*T2onlyonereasonfordephasing:PureT2dephasingisintrinsictosample(e.g.differentT2ofCSFandfat).T2*dephasingincludestrueT2aswellasfieldinhomogeneity(T2m)andtissuesusceptibility(T2ms).Duetotheseartifacts,Larmorfrequencyvariesbetweenlocations.T2*leadstorapidlossofsignal:imageswithlongTEwithhavelittlecoherentsignal.0.2SignalTE(s)001T2T2*14SusceptibilityartifactsMagnetfieldsinteractwithmaterial.Ferromagnetic(iron,nickel,cobalt)Stronglyattracted:dramaticallyincreasesmagneticfield.allsteelhasIron(FE),butnotallsteelisferromagnetic(tryputtingamagnetonaausteniticstainlesssteelfridge).Paramagnetic(Gd)Weaklyattracted:slightlyincreasesfield.Diamagnetic(H2O)Weaklyrepelled:slightlydecreasesfield.15TissueSusceptibilityDuetospin-spininteractions,hydrogen’sresonancefrequencydiffersbetweenmaterials.E.G.hydrogeninwaterandfatresonateatslightlydifferentfrequencies(~220Hz;1.5T).Macroscopically:Theseeffectscanleadspatialdistortion(e.g.‘fatshift’relativetowater)andsignaldropout.Microscopically:fieldgradientsatboundariesofdifferenttissuescausesdephasingandsignalloss.16FieldInhomogeneityArtifactsWhenweputanobject(likesomeone’shead)insideamagnet,thefieldbecomesnon-uniform.Whenthefieldisinhomogeneous,wewillgetartifacts:resonancefrequencywillvaryacrossimage.Priortoourfirstscans,thescanneris‘shimmed’tomakethefieldasuniformaspossible.Shimmingisdifficultnearair-tissueboundaries(e.g.,sinuses).Shimmingartifactsmoreintenseathigherfields.17SpinEchoSequenceSpinechosequencesapplya180orefocusingpulsehalfwaybetweeninitial90opulseandmeasurement.Thispulseeliminatesphasedifferencesduetoartifacts,allowingmeasurementofpureT2.Spinechodramaticallyincreasessignal.SignalTime01T2T2*0.5TE0.5TEActualSignal18SpinEchoSequencesTherefocusingpulseallowsustorecovertrueT2.ImagefromWebsiteincludesinteractiveadjustmentofT1/T2T2T2*19AnalogyforSpinEchoConsidertwoclocks.Clock1:minutehandtakes70minutestomakearevolution.Clock2:minutehandtakes55minutestomakearevolution.Simultaneously,setbothclockstoread12:00.(~sendin90oRFpulse).WaitpreciselyonehourMinutehandsnowdiffer:outofphase.Reversedirectionofeachclock(~sendin180oRFpulse).WaitpreciselyonehourMinutehandsnowidentical:bothreadnoon.TheyarebrieflybackinphaseMinutehandrotation0420o1hour1hour20T2*:fMRISignalisanartifactfMRIis‘BloodOxygenationLevelDependent’measure(BOLD).Brainregionsbecomeoxygenrichafteractivity:ratioofHbr/HbrO2decreases21BOLDeffectDeoxyhemoglobin(Hbr)actsascontrastagentFrequencyspreadcausessignallossovertimeEffectincreaseswithdelay(TE=echotime)But,overallsignalreduceswithTE.OptimalBOLDTE~60msfor1.5T,~30msat3T.Feraetal.(2023)JMRI19,19-26www.fmrib.ox.ac.uk/~karla/LowHighFrequency0.2TE(s)022BOLDartifactsfMRIisaT2*image–wewillhavealltheartifactsthataspin-echosequenceattemptstoremove.Dephasingnearair-tissueboundaries(e.g.,sinuses)resultsinsignaldropout.BOLDNon-BOLDwww.fmrib.ox.ac.uk/~karla/23OptimalfMRIscansMoreobservationswithshorterTR,butslightlylesssignalperobservation(duetoT1effectsandtemporalautocorrelation).WhenyouhaveasingleanatomicalregionofinterestusethefewestslicesrequiredforaveryshortTR.Forexploratorygroupstudy,useascanthatcoverswholebrainwithminimalspatialdistortion(forgoodnormalization).Typical3T:3x3x3mm64x64matrix,36slices,SENSEr=2,TE=35ms,TR=2100msTypical1.5T:3x3x3mm64x64matrix,36slcies,TE=60ms,TR=3500ms.ShorterTRyieldsbetterSNRDiminishingreturnsG.H.Glover(1999)‘OnSignaltoNoiseRatioTradeoffsinfMRI’24DiffusionImagingDiffusionimagingisanendogenouscontrast.Applytwogradientssequentiallywithoppositepolarity.Stationarytissuewillbebothdephasedandrephased,whilespinsthathavemovedwillbedephased.Sensitivetoacutestroke(DWI,seelesionlecture)Multipledirectionscanmeasurewhitematterintegrity(diffusiontensorimaging,seeDTIlecture)waterdiffusesfasterinunconstrainedventriclesthaninwhitematter25GadoliniumEnhancementGdPravenously-injected.GdnotdetectedbyMRI(1H).Gdhasaneffectonsurrounding1H.GdshortensT1,T2,T2*ofsurroundingtissue.makesvessels,highlyvasculartissues,andareasofbloodleakageappearbrighter.Veryraresideeffect:allergicreaction.Gdcanhelpmeasureperfusion.Usefulforclinicalstudies:howmuchbloodisgettingtoaregion,howlongdoesittaketogetthere?26TimeofFlightToFisamotioncontrast.InT1scans,motionofbloodbetweenslicescancauseartifacts.ToFintentionallymagnifiesflowartifacts.SeveralProtocolsofToF,E.G:UseveryshortTR,sosignalinsliceissaturated.Externalspinsflowingintoslicehavefullmagnetization.ConductaSpinEchoScan:90oand180oinversionpulsesappliedtodifferentslices.Onlynucleithattravelbetweenslicesshowcoherentsignal.SaturatedSpinsUnsaturatedSpinsSLICEFlow27ArterialSpinLabellingASLisanexampleofamotioncontrastIMAGEperfusion=IMAGEuninverted–IMAGEinvertedPerfusionisusefulforclinicalstudies:howmuchbloodisgettingtoaregion,howlongdoesittaketogetthere?www.fmrib.ox.ac.uk/~karla/inversionslabimagingplaneexcitationinversionxyz(=B0)bloodwhitematt