拜耳ALYCE：利用数据智能推进药物开发.pdf

1、/BAYERS ALYCE:ADVANCING DRUG DEVELOPMENT WITH DATA INTELLIGENCEDirk HelwigCDS&AClinical Data Sciences&AnalyticsJune 2024Decentralized Clinical TrialsData from Digital Health Technologies collected at optimal time points e.g.:Actigraphy devicesPhysical Activity Energy ExpenditureSleep Activity Body P

2、osture ECG Patches Heart RateRespiratory Rate R-to-R IntervalFluid StatusData collected at on-site visitsDemographicsWeight,HeightVital SignsSingle 12 Lead ECGsAdverse Events,Medical HistoryVolume,Speed and Complexityof generated data will increase2/DATA+AI SUMMIT /June 20243Operational database(Cli

3、nical Data Warehouse)Sleep dataeCOAActivity data other data sourcesIngestionQuality checkMerge,WranglingVisualizationsNotificationsEnd-2-End PipeliningAdvanced Analytics231Data ingestionData orchestration,Reports/Dashboards&Advanced AnalyticsECG dataPush of Processed DataIngesting/Wrangling/Analysin

4、g/DATA+AI SUMMIT /June 2024Therapeutic Area Overarching(TAO)dashboards(AE)4/DATA+AI SUMMIT /June 2024Therapeutic Area Overarching(TAO)dashboards(Labs)5/DATA+AI SUMMIT /June 2024Clinical Trial Anomaly Spotter(CTAS)6CTAS(Clinical Trial Anomaly Spotter)is a CSM(Central Statistical Monitoring)tool focus

5、ed on identifying anomalous time series of continuous variables.“Unsupervised”tool which does not use pre-specified KRIs.The identified anomaliescan be such that were not anticipated during study design.Originally developed by Bayer where it is available for all company studies.The Bayer-internal to

6、ol was presented in PHUSE EU Connect 23.Please see refs.1 and 2 for details.Code base shared with IMPALA for co-development in July 2023.CTASClinical trialdata-Parameter selection-Data reformatting-Definition of specific time seriesData pre-processing-Visualization-Signal review-Signal trackingResul