Skip to main content
Grant Details

Grant Number: 1UG3CA225021-01 Interpret this number
Primary Investigator: Saltz, Joel
Organization: State University New York Stony Brook
Project Title: Methods and Tools for Integrating Pathomics Data Into Cancer Registries
Fiscal Year: 2018
Back to top


Abstract

ThegoalofthisprojectistoenrichSEERregistrydatawithhigh?qualitypopulation?based biospecimendataintheformofdigitalpathology,machinelearningbasedclassificationsand quantitativepathomicsfeaturesets.Wewillcreateawell?curatedrepositoryofhigh?quality digitizedpathologyimagesforsubjectswhosedataisbeingcollectedbytheregistries.These imageswillbeprocessedtoextractcomputationalfeaturesandestablishdeeplinkageswith registrydata,thusenablingthecreationofinformation?rich,populationcohortscontaining objectiveimagingandclinicalattributes.SpecificexamplesofdigitalPathologyderivedfeature setsincludequantificationoftumorinfiltratinglymphocytesandsegmentationand characterizationofcancerorstromalnuclei.Featureswillalsoincludespectralandspatial signaturesoftheunderlyingpathology.Thescientificpremiseforthisapproachstemsfrom increasingevidencethatinformationextractedfromdigitizedpathologyimages (pathomicfeatures)areaquantitativesurrogateofwhatisdescribedinapathologyreport.The importantdistinctionbeingthatthesefeaturesarequantitativeandreproducible,unlikehuman observationsthatarehighlyqualitativeandsubjecttoahighdegreeofinter?andintra?observer variability.Thisdatasetwillprovide,aunique,population?widetissuebasedviewofcancer, anddramaticallyaccelerateourunderstandingofthestagesofdiseaseprogression,cancer outcomes,andpredictandassesstherapeuticeffectiveness. ThisworkwillbecarriedoutincollaborationwiththreeSEERregistries.Wewillpartner withTheNewJerseyStateCancerRegistryduringthedevelopmentphaseoftheproject(UG3). Duringthevalidationphaseoftheproject(UH3),theGeorgiaandKentuckyStateCancer Registrieswilljointheproject.Theinfrastructurewillbedevelopedinclosecollaborationwith SEERregistriestoensureconsistencywithregistryprocesses,scalabilityandabilitysupport creationofpopulationcohortsthatspanmultipleregistries.Wewilldeployvisualanalytictools tofacilitatethecreationofpopulationcohortsforepidemiologicalstudies,toolstosupport visualizationoffeatureclustersandrelatedwhole?slideimageswhileprovidingadvanced algorithmsforconductingcontentbasedimageretrieval.Thescientificvalidationofthe proposedenvironmentwillbeundertakenthroughthreestudiesinProstateCancer,Lymphoma andNSCLC,ledbyinvestigatorsatthethreesites.

Back to top


Publications

Adversarial Domain Adaptation for Classification of Prostate Histopathology Whole-Slide Images.
Authors: Ren J. , Hacihaliloglu I. , Singer E.A. , Foran D.J. , Qi X. .
Source: Medical Image Computing And Computer-assisted Intervention : Miccai ... International Conference On Medical Image Computing And Computer-assisted Intervention, 2018 Sep; 11071, p. 201-209.
EPub date: 2018-09-26 00:00:00.0.
PMID: 30465047
Related Citations




Back to Top