RFID(医疗应用)_Application_on_medication.ppt

1 RFIDapplication system 69521001陳盈君69521038吳哲維69521040林冠良 2 outline IntroductionDrugSafetyUsingRFID 3 Introduction RFID RadioFrequencyIdentificationSomeRFIDapplication 4 RFIDapplication WewillproposehowusingRFIDinhospital DrugSafetySmartHospitalOperatingRoom 5 DrugSafety TheapplicationofRFIDondrugsafetyofinpatientnursinghealthcare 6 DrugSafety Reducingmedicalerrorsistheprimarygoaltoenhancepatientsafety ThebottleneckofreductionisthattheHospitalInformationSystem HIS cannottracktheinformationaboutthepatientname operationaltimeandlocationimmediatelyandcorrectly 7 MedicalMistake Errorsofomission Itmeansthatthephysicianforgetssomedrugsorprescribesinappropriatedosageintheprescription Errorsofcommission Itconcernsaboutthosewrongdrugsaresuppliedtowrongpatientsinthewrongtime 8 UsingRFIDonDrugSafety TheRFIDisapowerfultoolinmanagingthehospital WiththeattachmentofRFIDondrugs themedicinestaffcanexactlyidentifythedrugsandknowtheirprecisepositionandmovement Thus themedicinestaffcaneffectivelyavoidthemisusageofmedicineontothepatient 9 Systemenvironment TheresearchincludesthetechnologicaldocumentsthatRFIDtechnicalcommitteeonEPCGlobal Basedonthe915Mhzpassiveread writeRFIDtagandbarcodeco system andHL7 XML SOAPdataexchangestandardinsidethehospital 10 SystemDesign Somesubsystem function MedicinetagsubsystemSavantmanagementsubsystemMedicinecomparisonsubsystemMedicineentryandexitsubsystemPatiententryandexitsubsystem 11 Medicinetagsubsystem PrintthetagsManagetheinputofrelevantmaterialsofmedicinesmainly 12 Savantmanagementsubsystem DealwithRFIDrelevantequipment Itgathersdatafromtagsandsendsthedatarevisiontoreaderdevicethatcarryouttherevisionintag ReviseasetofmaterialsofRFIDringonpatients InputandrevisetheRFIDtagmaterialsondrug RevisethesetupofRFIDreaderdeviceandreceiveandfilterRFIDmaterials 13 Medicinecomparisonsubsystem Dealwithconfirmingrelevantpatientmedicinemainly Examinepatient smedicineandpatient srecordUsethemedicinecomparisonfunction 14 Medicineentryandexitcontrol Proceedingthedrugentryandtransferouttotheareamainly Themedicinecanenterallowedareasofthereaderdevicechosen 15 Patiententryandexitcontrol Dealwithpatient srouteofmovingmainly Itcanrestrainthepatientfromenteringlimitedarea orrestrainthepatienttoleavewardinmedicinetime 16 SystemFunction 17 Systemstructure Wecandividethissysteminto4parts TagproducingRFIDmanagementDrugsafetyPatientsafety 18 Tagproducing Afterthedoctormaketheorderviathedoctor sordersystem thedruggiststickRFIDtagondrugviadrugstagsystem DrugRFIDtagsPatientRFIDhandring 19 RFIDmanagement Itincludesavantmanagementsubsystem hand holdreaderandfixedreaderdevice Thehand holdreaderdevicesareprovidedtorelevantnursingstaffmainlyandtheycancheckdrugsanddosage AndtodetectandexamineRFIDhandringintheareathroughfixedreaderdevice thesystemnotesdownforpatientcontacthistoryandflowofthedrugs 20 Drugsafety Includemedicinecomparisonandmedicineentryexitcontrolsubsystem 21 Patientsafety Includepatiententryexitcontrolsubsystem Thisdetectsexaminingtopatient sflowmainly 22 Systemarchitecture 23 Systemcorrelation 24 WhereProducingStreamingData Thefixedreaderreceivethestreamingdata becausethetagsandringsarealwaysexistinhospital Entryexitcontrolsystemwillsendstreamingdatatosavantsystem Hand holdreaderalsosendsstreamingdatatosavantsystemAtlast thesavantsystemalsosendstreamingdatatocomparisonsystem 25 eWellness BuildingaSmartHospitalbyLeveragingRFIDNetworks 26 eWellness Introduction 1 3 TheprimaryapplicationwouldbetheuseofembeddedRFIDtagsinidentificationbadgesforhealthcareprofessionals patients andfixedassets Thusallowstheabilitiesto 1 trackingdata 2 off lineanalysis 3 drugdispensation 分配 andvalidatemedicaldosages 劑量 4 improvedaccuracyoflaboratoryresults 27 eWellness Introduction 2 3 However thelargevolumesofdatageneratedfromRFIDnetworksresultsinaimpedancematchingproblem Thedatacanbegathered thecoreofchallengingproblemsresidesintheincreasinglycomplexdatamanagementandinterpretationissues Forexample Passingfilteredmeaningfuldatatohospitalenterpriseapplications managingsensordataattheedgeofthenetworkwhereRFIDtagsandsensornodesarelocated andcorrectlyinterpretingaggregatedsensordatainthecontextoftheapplicationdomains 28 eWellness Introduction 3 3 Weproposeanevent basedarchitecturalsolution calledeWellness 29 Event BasedInformationIntegrationParadigm EventDefinition Givenkevents1 k theithevent ei eventidt eventtemporallys spatiallocation attributes suchaseventname eventtopic 30 SystemArchitectureOfeWellness SystemArchitecture 31 SystemArchitectureOfeWellness Real timeeventcollectionandfilteringmodule Characterstringscanprovideatextualrepresentationoffilterconstraints Theeventfilterclassrepresentedbyobjectcarriesoutfilteringfunctionsbyexecutingitsfunctions Forunderstandingsemantics Toensurefilteringeventsinreal time 32 SystemArchitectureOfeWellness Eventhierarchicalmodel basiceventsthatcanbecomputeddirectlyfromsensorinput mappingdataeventsreliesontheeventrelationshipsdetectedinsensordatastreams interpretstheinteractionsofentities primitiveeventsinabusinesscontext 33 SystemArchitectureOfeWellness Semanticrepresentationmodule SemanticrepresentationofthedynamicdatastreamscapturedfromsensorandRFIDnetworkscanbeuseful forexample inreferringrelationshipsbetweenevents AnEventOntology EO isproposed ItisexpressedbythestandardOWL WebOntologyLanguage andextendsandintegratesontologiesfortime space andevents 34 SystemArchitectureOfeWellness Semanticrepresentationmodule EOisgroupedintofourdistinctiveontologydocumentsasfollowing 1 Theontologydefinestypicalvocabulariesfordescribinganobjectwithinanevent 2 Itrepresenttheinstantandintervaltime e qinstant tme before tme after tme beforeOrAt tme afterOrAt andtme sameTimeAs 3 Defineasymbolicrepresentationofspaceandsupportreasoningaboutthespatialrelationsbetweenvarioustypesofgeographicalregions 4 Describetheoccurrenceofdifferentactivities schedules andsensingevents e qeve Eventclassrepresentsasetofalleventsinthedomain 35 SystemArchitectureOfeWellness Semanticrepresentationmoduleexample ThefollowingexampleshowshowEOcanbeusedtodescribeaneventinwhichanoperationalinstrumenthasbeendetectedunderworkfrom2005 04 28at10 30 21ESTto2005 04 28at11 40 00EST andtheeventoccursatalocationthatisdescribedbytheroomnumberandbuildingnumber 36 SystemArchitectureOfeWellness Semanticrepresentationmodule 37 APervasiveComputingSystemfortheOperatingRoomoftheFuture AprototypeContextAwarePerioperativeInformationSystem Introduction ThisworksupportedinpartbyasubcontractfromUMMedicalSchoolincontextoftheTraumapodProjectWedescribeaprototypeContextAwarePerioperativeInformationSystemtocaptureandinterpretdatainanoperatingroomofthefuture Thecaptureddataisusedtoconstructthecontextofthesurgicalprocedureanddetectmedicallysignificantevents Suchevents andotherstateinformation areusedtoautomaticallyconstructanElectronicMedicalEncounterRecord EMR電子病歷 38 Introduction Motivation Performingasurgeryisanelaborateprocessandproceedsinprogressivestages Forexample thepreoperativephaseofasurgeryinvolvesidentifyingthepatient determiningreadinessofthenursingandmedicalstaff preparingtheoperatingroom andcapturingdatafromincomingmedicalrecordsregardingvitals pre operativemedications tests andscansetc 39 Introduction Motivation Goal Somefractionofthesepatientsareunscheduledandidentiedonlyonthedayofsurgery thisfractioncanbelargeinfacilitiesthatdealwithtrauma Theresultingchaoscanbeoverwhelming evenwithsomeformofelectronichealthrecord EHR system Sowewishthesystemmustkeeptrackof andarchive whatisdonetoeachpatient bothdiagnostically 診斷的 andtherapeutically 治療的 how when andwhyitisaccomplished andhowthepatientresponds 40 Introduction Goal cont OurgoalistodevelopaContext AwarePerioperativeInformationSystemthatwillautomatemostoftheperioperativesupportfunctionssuchaspatienttracking inventorymanagement clinicaldocumentationetcwiththehelpofpervasivecomputingandsemanticwebtechnologies 41 Introduction Goalenvironment Weseektocreateasystemthatcanworkintwoenvironments Oneisthe OperatingRoomoftheFuture ThistermisusedtodescribetheconceptthatissoughttoberealizedbyseveralsurgicalgroupsacrossthecountryandinvolvesanITenabledOperatingRoom ThesecondisTraumaPod whoseaimistodevelopanautomatedsurgicaltreatmentsystemthatdoesnotrequireonsitemedicalpersonnelonthefrontlinesofbattle 42 Introduction FirstStep Thefirstphaseoftheprogramisanefforttodeveloprobotictechnologytoperformatotallyunmannedsurgicalprocedurewithinafixedfacilityinanormalsurgicalsetting Ahumansurgeonwillconductalltherequiredsurgicalproceduresfromaremotelocationusingasystemofsurgicalmanipulators 43 Background Datacharacteristic Clinicalrecordkeepinginhighvelocityhealthcaredeliveryenvironmentslikesurgeryisanecessaryandcriticaltask Itisalsoatimeconsumingtaskthatdetractsfromhandsonpatientcareandcontributestoextraordinarylaborcostsassociatedwithcollecting transcribingandre keyingrecordsthroughouttheperioperativeprocess 44 Background Datacollection Datacollectionintheoperatingroomiscomplicated Dueto 1 multipleprovidersrecorddataforasinglecareevent 2 informationcollectedbyoneproviderisnotreadilyavailabletoanother 3 peoplewithdifferentexperiencesneedvariousdata 45 Background Dataaccuracy AnElectronicMedicalEncounterRecord EMR hasthepotentialtoreducedocumentationerrorsbyminimizingdataredundancyandprovidingaccuratedetailsoftheongoingsurgery 46 Background Mostsensingtechnologydetectslowlevelevents Sowhatismorevaluableistheknowledgeatagiventime Wedefineamedicallysignificanteventasanyeventthataffectsorisapartofthesurgicalprocedure InourresearchweusetheRadioFrequencyIdentification RFID systemtodetectmedicalsupplies toolsandthestaff 47 SystemArchitecture Theframeworkwedevelopedisdesignedtocollect processandreasonoversurgicalcontextinformationtodetectmedicallysignificantevents 48 SystemArchitecture Itconsistsofa3 tiereventdetectionsystem 49 dataiscollectedfromvarioussensorsintheoperatingroomandprocessedbyastreamprocessortoidentifylowleveleventslikepresence absenceofanRFIDtag detectlowlevelmedicaleventssuchashighbloodpressure apneaetc Eventdetectionatthislevelalsomakesuseofpatientsmedicalhistoryandknowledgeofmedicinesusedtodetectevents therulebasedenginethatcorrelateseventsfromthesecondtiertodetectmedicallysignificanteventssuchasadministrationofanesthesia tensionpneumothoraxetc SystemArchitecture Datasource PatientMonitorsAssetandPersonnelTrackingSystemsMessageExchangesWheretheoperatingroomishighlyautomated asintheTraumapodsystem ForexampleinTraumapod wecanlistentoexchangesbetweenthe ToolRacksystem andthe scrubnursesystem tofigureoutwhichtoolshavebeenmovedfromtheracktothe robotic nurse Thiscanbedonewithoutanyexplicitsensingtechnologythatwouldbeneededinaliveoperatingroomwithhumansurgeonsandnurses 50 SystemArchitecture Datasourceinoperatingroom 51 52 APervasiveComputingSystemfortheOperatingRoomoftheFuture part2 53 DetectingMedicallySignificantEvents Thesystemdefinetherulestodetectthemedicallysignificantevents Theknowledgebasewasdevelopedbygatheringinformationfromananesthesiologistbyinterviewsandfrommedicalliteraturewhichdescribemethodstoanalyzeandinterpretphysiologicaldata 54 DetectingMedicallySignificantEvents Physiologicalparametersreflectapatients healthstatus Interpretationofphysiologicaldatatoinferthepatientsconditionisachallengeproblem Systemshavehighrateoffalsepositiveorfalsenegativeevents Reason Physiologicalparametersnotonlydependonthephysiologicprocessesbutalsoonfactorssuchaspatientscurrentcondition medicalhistory medicinesadministeredandthesequenceofoccurrenceofotherevents Mostsystemsconsideronlyasubsetofthesefactors 55 DetectingMedicallySignificantEvents EX Earliestdataanalysissystemsusedsimplelimitsonphysiologicalparametersforbasicinterpretation Eachphysiologicalparameterhasarangeofvaluesthatcanbeclassifiedasnormalorabnormal Interpretationofaparametervarieswiththeclinicalcontext Thereisnosetthresholdthatwillclassifythevalueasnormalorabnormal usingthefuzzysettheorytocapturethisuncertaintyinmedicaldata fuzzymembershipfunctionsareusedtoclassifydatavalues thevaluecanbeverylow low normal highorveryhigh 56 DetectingMedicallySignificantEvents rateofchangecanbeconstant stableorabrupt andthevaluecanbeincreasing ordecreasing thevalueandrateofchangeofvalueisusedtodetectevents Thesetpointsusedtodefinetherangeofvaluesvarieswitheachpatient EX therangeofnormalbloodpressureforahypotensivepatientwillbedifferentthantherangeforapatientwithnormalbloodpressure Future systemscansettheselimitsbyanalyzingthepatientsmedicalhistoryandthepre opdiagnosis 57 DetectingMedicallySignificantEvents Giventhedatavalueofaphysiologicalparameteranditstrend thetechniquesweusedtocorrelatelowleveleventsare Pre opDiagnosisMulti variableAnalysisEventHistoryEffectofMedicines 58 DetectingMedicallySignificantEvents Pre opDiagnosis beforethepatientisbroughtintothesurgery thepatient sconditionisevaluated includingtakingnoteofthevitalsigns anymedicalcareprovided aphysicalexaminationandanyothernotablemedicalcondition EX ifthepatientwasbleedingexcessivelypriortothesurgery detectingfluidinfusionsduringthesurgeryismoreaccurate 59 DetectingMedicallySignificantEvents Multi variableAnalysis physiologicparametersnotonlydependonphysiologicprocessesbutarealsoaffectedbythepatientscurrentcondition EX lowanddecreasingbloodpressure doesnotgivesignifytoomuchdetail However lowanddecreasingbloodpressurewithhighandincreasingheartrateimpliespotentiallossoffluids Monitoringaphysiologicalparameteralongwithitsrelationshipwithotherparametershelpsdeterminemoremeaningfulevents 60 DetectingMedicallySignificantEvents EventHistory Eventhistoryisasetoflow levelandhigh leveleventsalreadydetected Ahighleveleventisoftenacompositionoflow leveleventsandpotentiallyotherhighlevelevents Example EventConjunctionIf TensionPneumothorax and SystolicBP low and increasing and HeartRate high and decreasing thenEvent Decompression 61 DetectingMedicallySignificantEvents EffectofMedicines Themedicinesadministeredduringthesurgery mayormaynothaveasignificanteffectonthepatient sphysiology Thetimethemedicinewasadministered effectsexpected durationofeffectandtimetoaffectaresomeofthefactorsthatneedtobetakenintoaccounttodetecttheireffectinthephysiologicalparameters Determiningtheeventofmedicineadministrationisextremelydifficult Solution themedicalsuppliesaretaggedusingRFIDtags wecanacquirethecontextualinformationusingwhichwecandetectmedicineadministration 62 DetectingMedicallySignificantEvents Forthosemedicineswhoseeffectisobservedinthephysiologicalparameters aneventissignaledtoindicatethatthemedicinewasactuallyused EX If RespiratoryRate abruptdecrease andSystolicBP abruptdecrease andHeartRate stable If TimeAnestheticdetected 40sec thenEvent StartAnesthesia 63 SystemInterface ElectronicMedicalRecord TheEMRcanbedisplayedonacomputerscreenintheoperatingroom Itprovidesasummaryofthepatientprofile thepre opdiagnosis laboratoryreports andradiology imagingreports Ability Beforesurgerycancommence obviouserrorssuchasrightpatient rightsurgeon rightequipmentandthepresenceofconsentform assumedtobeRFIDtagged arechecked Duringsurgery thevitalsignsofthepatientarestreamedtothescreenduringthesurgery postsurgery videoclipsofdefinedlengthforeachoftheclinicallysignificanteventsarecreatedandthecorrespondingvideoURLisstoredinthemedicalencounterrecord 64 65 Ev