Supporting Depth and Motion Perception in Medical Volume Dat

There are many application areas where dynamic visualization techniques cannot be used and the user can only view a still image. Perceiving depth and understanding spatio-temporal relations from a single still image are challenging tasks. We present visual

SupportingDepthandMotionPerceptioninMedicalVolumeData

JennisMeyer-Spradow,TimoRopinski,andKlausHinrichs

VisualizationandComputerGraphicsResearchGroup,DepartmentofComputerScience,UniversityofM¨unster

{spradow,ropinski,khh}@math.uni-muenster.de

Therearemanyapplicationareaswheredynamicvisualizationtechniquescan-notbeusedandtheusercanonlyviewastillimage.Perceivingdepthandunderstandingspatio-temporalrelationsfromasinglestillimagearechal-lengingtasks.Wepresentvisualizationtechniqueswhichsupporttheuserinperceivingdepthinformationfrom3Dangiographyimages,andtechniqueswhichdepictmotioninherentintime-varyingmedicalvolumedatasets.Inbothcasesnodynamicvisualizationisrequired.

1Introduction

Volumerenderinghasbecomeamature eldofinteractive3Dcomputergraph-ics.Itsupportsprofessionalsfromdi erentdomainswhenexploringvolumedatasets,representingforexamplemedicalormeteorologicstructuresandprocesses.Currentmedicalscannersproducedatasetshavingahighspatialresolution,andevendynamicdatasetscanbeobtained.EspeciallyComputerTomography(CT)datasetsarehighlysuitableforshowinganatomicalstruc-tures.However,whendynamicprocesseshavetobeexploredasingle3Dvol-umedatasetisofteninsu cient.Forexample,PositronEmissionTomography(PET)imagingtechniquesareusedforexploringthedynamicsofmetabolism.Successivescansproduceatimeseriesofimageseachshowingthedistribu-tionofmoleculesatacertainpointintime.Time-varyingvolumedataisalsoproducedwhenapplyingfunctionalMagneticResonance(MR)imagingtechniquestoevaluateneurologicalreactiontocertainstimuli.

Although3Dvolumescanbedisplayedseparatelytoviewthedataforspeci cpointsintime,thedynamicscontainedinatime-varyingdatasetcanonlybeextractedwhenconsideringallofthem.Thereforetime-varyingvol-umedatasetsareoftenvisualizedbydisplayingthe3Dvolumessequentiallyintheordertheyhavebeenacquired.Whenrenderingthese3Dvolumesatahighframerate,theviewerisabletoconstructamentalimageofthedy-namics.Besidesthepossibilitytoacquiretime-dependentdataalsothespatial

There are many application areas where dynamic visualization techniques cannot be used and the user can only view a still image. Perceiving depth and understanding spatio-temporal relations from a single still image are challenging tasks. We present visual

2JennisMeyer-Spradow,TimoRopinski,andKlausHinrichs

resolutionofmedicalscannershasbeenincreasedleadingtolargeandpossiblymorecomplexdatasets.Thiscomplexitydemandsthedevelopmentofinter-activevisualizationtechniquessupportingane cientande ectiveanalysis.Sincethemedicaldatasetsacquiredthroughdi erentimagingtechnologiesdi erinthewaytheyareexplored,specializedtechniqueshavetobede-veloped.Weproposevisualizationtechniqueswhichhavebeendevelopedtosupporttheexplorationofangiographydatasets.Wewillshowhowthedepthperceptionandthusthespatialcognitionofthesedatasetscanbeimprovedandthusresultsinamoree cientaswellase ectivediagnosis.

Oftendynamicsareusedtovisualizebothtime-varyingaswellashighres-olutiondatasets.Inthelattercaserotationsorotherinteractivetechniquesareappliedtoexplorethedatasets.However,insomedomainsthesetechniquesareinappropriate,sinceviewerspreferstillimagesratherthandynamicim-agesequences.Thefollowingtwoproblemsarisewhendealingwithimagesequencesof3Dvolumes.Whenseveralpeoplewatchanimagesequence,aviewerusuallypointsataspeci cfeaturewithinanimagetoexchange nd-ingswiththeotherparticipants.Sincetheimagesaredynamic,itishardtopinpointsuchanitemofinterest,especiallyifitismoving,andtheanima-tionneedstobepaused.Anotherproblemistheexchangeofvisualizationresults.Inmedicaldepartmentsdiagnostic ndingsareoftenexchangedonstaticmediasuchas lmorpaper.However,itisnotpossibletoexchangetime-varyingvolumedatasetsinthismanner.Althoughthesequentialview-ingprocesscouldbesimulatedbyshowingall3Dvolumesnexttoeachother,thiscouldleadtoregistrationproblems,i.e.,itwouldbemoredi culttoidentifyareferenceitemacrossthe3Dvolumes.

Weaddresstheseproblemsbyproposingvisualizationtechniqueswhichrepresentthemotiondynamicsextractedfromatimeseriesof3Dvolumeswithinasinglestaticimageandsupporttheexplorationofhighresolutiondatasets.AllpresentedtechniquesareimplementedinourVolumeRenderingEngineVoreen[Vor07].

2RelatedWork

DepictingMotion

Mainlynon-photorealisticrenderingtechniquesareusedfordepictingmotioninasingleimage.Masuchetal.[MSR99]presentanapproachtovisualizemotionextractedfrompolygonaldatabyspeedlines,repeatedlydrawncon-toursandarrows;di erentlinestylesandotherstylizationgivetheviewertheimpressionofahand-drawing.NienhausandD¨ollner[ND05]presentatech-niquetoextractmotioninformationfromabehaviorgraph,whichrepresentseventsandanimationprocesses.Theirsystemautomaticallycreatescartoon-likegraphicalrepresentations.JoshiandRheingans[JR05]usespeedlinesand owribbonstovisualizemotioninvolumedatasets.However,theirapproach

There are many application areas where dynamic visualization techniques cannot be used and the user can only view a still image. Perceiving depth and understanding spatio-temporal relations from a single still image are challenging tasks. We present visual

SupportingDepthandMotionPerceptioninMedicalVolumeData3handlesonlyfeatureextractiondataandismainlydemonstratedonexperi-mentaldatasets.Meyer-Spradowetal.[MRVH06]usespeedlinestovisualizemotiondynamicsextractedinapre-processingstep.Furthermoretheyuseedgedetectionandtransparencytechniquestoshowseveraltimestepsinonestillimage.

Alotofresearchintheareaofdynamicvolumevisualizationhasbeendedicatedto owvisualization.Motionarrowsarewidelyused[LHD*04],butareoftendi culttounderstandbecauseinformationislostwhenperforminga2Dprojectiontothescreen.BoringandPang[BP96]triedtotacklethisproblembyhighlightingarrows,whichpoin …… 此处隐藏：28154字，全部文档内容请下载后查看。喜欢就下载吧 ……