2022
Robust convolutional neural networks against adversarial attacks on medical images
Shi X, Peng Y, Chen Q, Keenan T, Thavikulwat A, Lee S, Tang Y, Chew E, Summers R, Lu Z. Robust convolutional neural networks against adversarial attacks on medical images. Pattern Recognition 2022, 132: 108923. DOI: 10.1016/j.patcog.2022.108923.Peer-Reviewed Original ResearchConvolutional neural networkMedical imagesAdversarial attacksAdversarial perturbationsNeural networkRobust convolutional neural networkNovel defense methodMedical image modalitiesReal-world scenariosSignificant security risksDefense methodsFeature representationSecurity risksHuman expertsNoisy featuresAttacking methodImage modalitiesAttacksImagesNetworkMedical applicationsOriginal performanceSparsityPerformance deteriorationApplicationsMulti-task deep learning-based survival analysis on the prognosis of late AMD using the longitudinal data in AREDS.
Ghahramani G, Brendel M, Lin M, Chen Q, Keenan T, Chen K, Chew E, Lu Z, Peng Y, Wang F. Multi-task deep learning-based survival analysis on the prognosis of late AMD using the longitudinal data in AREDS. AMIA Annual Symposium Proceedings 2022, 2021: 506-515. PMID: 35308963.Peer-Reviewed Original ResearchConceptsAge-related macular degenerationImage featuresMulti-task learning frameworkConvolutional neural networkVision lossLate age-related macular degenerationEye Disease StudyLearning frameworkNeural networkFundus photographsPatient riskMacular degenerationStandard featuresSevere formComplex featuresSurvival analysisCurrent visitLongitudinal dataDisease StudyHistorical dataRapid paceFeaturesNetworkAREDSPatients
2021
Benchmarking Effectiveness and Efficiency of Deep Learning Models for Semantic Textual Similarity in the Clinical Domain: Validation Study
Chen Q, Rankine A, Peng Y, Aghaarabi E, Lu Z. Benchmarking Effectiveness and Efficiency of Deep Learning Models for Semantic Textual Similarity in the Clinical Domain: Validation Study. JMIR Medical Informatics 2021, 9: e27386. PMID: 34967748, PMCID: PMC8759018, DOI: 10.2196/27386.Peer-Reviewed Original ResearchSemantic textual similarityConvolutional neural networkDeep learning modelsReal-time applicationsDL modelsSentence pairsNeural networkTextual similarityBERT modelNational Natural Language Processing Clinical ChallengesLearning modelNatural language processingAverage Pearson correlationData setsDifferent similarity levelsInference timeGeneralization capabilityManual annotationLanguage processingPearson correlationEnsemble modelWord orderTime efficiencyNegation termsTraining set