An open source automated tumor infiltrating lymphocyte algorithm for prognosis in melanoma
Acs B, Ahmed FS, Gupta S, Wong P, Gartrell RD, Sarin Pradhan J, Rizk EM, Gould Rothberg BE, Saenger YM, Rimm DL. An open source automated tumor infiltrating lymphocyte algorithm for prognosis in melanoma. Nature Communications 2019, 10: 5440. PMID: 31784511, PMCID: PMC6884485, DOI: 10.1038/s41467-019-13043-2.Peer-Reviewed Original ResearchConceptsOpen sourceOpen source softwareSource softwareTIL scoreTraining setDisease-specific overall survivalHigh TIL scorePoor prognosis cohortsSubset of patientsAlgorithmIndependent prognostic markerBroad adoptionAssessment of tumorOverall survivalFavorable prognosisMelanoma patientsMultivariable analysisValidation cohortIndependent associationPrognostic markerSeparate patientsPrognostic variablesIndependent cohortRetrospective collectionMelanomaPrediction of distant melanoma recurrence from primary tumor digital H&E images using deep learning.
Robinson E, Kulkarni P, Pradhan J, Gartrell R, Yang C, Rizk E, Acs B, Rohr B, Phelps R, Ferringer T, Horst B, Rimm D, Wang J, Saenger Y. Prediction of distant melanoma recurrence from primary tumor digital H&E images using deep learning. Journal Of Clinical Oncology 2019, 37: 9577-9577. DOI: 10.1200/jco.2019.37.15_suppl.9577.Peer-Reviewed Original ResearchDeep neural net architectureOpen source softwareRecurrent neural networkNeural net architectureDigital pathology toolsDeep learningSource softwareNet architectureFeature informationNeural networkNetwork parametersTIFF filesAdjuvant immunotherapyMelanoma recurrenceCohort 2Cohort 1Cell classificationStage IMultivariable Cox proportional hazards modelsDNNCox proportional hazards modelColumbia University Medical CenterNuclear segmentationEvidence of diseaseIndependent prognostic factor