資源描述：
《big data challenge – from emr’s to translational research分析課件》由會員上傳分享，免費在線閱讀，更多相關(guān)內(nèi)容在教育資源-天天文庫。
1、BigDataChallenge–FromEMR’stoTranslationalResearchMiroslavKon?ar,PhD OracleHealthcareBusinessDevelopmentDirector HL7CroatiaChair,AssociateProfessorforBiomedicalInformaticsWhatisBigDataWhyitisimportantforHealthcareProvidersNowMITBigDataDefinitionWhatkindofbi
2、garewetalkingabouthere?U.S.Healthcaredataisgrowingby15Petabytesaday–currentlyUpto80%ofhealthcaredataiscurrentlyunstructured.5millionbytes(megabytes)CompleteprintedworksofShakespeare20billionbytes(gigabytes)CompleterecordedworksofBeethoven10trillionbytes(te
3、rabytes)PrintedworksintheLibraryofCongress2quadrillionbytes(petabytes)GoogleEarthinformationmanagedtodayTypicalperson’sEHR~1MBhealthyyoung,noimages40MBmiddle-agedw/healthissues,noimages3-5GBwithseveralhealthissuesw/images600petabytes-10exabytes(exabyte=bil
4、liongigabytes)EstimatedsizeofcurrentUSdigitizedpatientdata1.5millionEstimateddifferentmedicaldevices(WHO)BigData:Themanagementanduseofultra-largeamountsofinformationManagementanduse=efficientstorage,search,analysis,visualizationUltra-large=>1PetabyteChroni
5、cDiseaseMgt.CarePatternsPopulationHealthPayforPerformanceTelemedicinePatientAdvocacyCustomerSatisfactionDiseaseMgt./PreventionSocialMediaInteractionsCare/ProcessImprovementsMarketing/BrandingWhereAreYourBigDataOpportunities?Lowercarecostandraisecarequality
6、CapacityManagementStaffingOptimizationProtocolOptimizationProcessEfficiencySupplyStandardizationCostAccountingPersonalPreferencesIndividualizedTreatmentBiometricsDataManagementApplyResearchNewClinicalProtocolsAccountableCarePatientEngagementPersonal&Transl
7、ationalMedicineOperationalExcellence1234WhyOracleforBigDataBigData:Oracle’sDefinitionLargevolumes,highchangevelocity,complexvariety,unknownvalueperbyteEmergingsourcesofdata,ie.sensor,machinegenerated,social,images,unstructured,etc.,thatexhibitanyorallof“th
8、efourV’s”,potentiallyintegratedwith“traditional”structureddataHealthcareproviderscanacquire,organize,process,analyzemassiveamountsofdata,primarilythroughtheuseofscale-outonlow-costhardware,non-relationaldatam
