Strategies to Improve the Robustness and Generalizability of Deep Learning Segmentation and Classification in Neuroimaging
Tran A, Zeevi T, Payabvash S. Strategies to Improve the Robustness and Generalizability of Deep Learning Segmentation and Classification in Neuroimaging. BioMedInformatics 2025, 5: 20. PMID: 40271381, PMCID: PMC12014193, DOI: 10.3390/biomedinformatics5020020.Peer-Reviewed Original ResearchDeep learning modelsGeneralizability of deep learning modelsLearning modelsArtificial intelligenceIntegration of deep learning modelsImprove model robustnessDomain shiftEnhance deep learningTransfer learningData augmentationDeep learningMedical imagesTask-specific applicationsModel robustnessSensitive to artifactsLearning segmentationDeep learning segmentationModel attributesNeuroimaging applicationsRobustnessClassificationDataUncertainty estimationLearningData variabilityDICOM LUT is a Key Step in Medical Image Preprocessing Towards AI Generalizability
Dapamede T, Li F, Khosravi B, Purkayastha S, Trivedi H, Gichoya J. DICOM LUT is a Key Step in Medical Image Preprocessing Towards AI Generalizability. Journal Of Imaging Informatics In Medicine 2025, 1-9. PMID: 39890738, DOI: 10.1007/s10278-025-01418-5.Peer-Reviewed Original ResearchDeep learning modelsHistogram equalizationInformation lossPreprocessing techniquesPre-processingLearning modelsPerformance of deep learning modelsMachine learning practitionersRisk of information lossDeep learning classifierImage preprocessing techniquesImprove model robustnessImage pre-processingTraining dataCXR datasetPreprocessed informationLearning classifiersDataset curationLearning practitionersModel performancePotential overfittingDatasetModel robustnessPreprocessingSharing datasetImproving the Robustness of Deep Learning Models in Predicting Hematoma Expansion from Admission Head CT.
Tran A, Abou Karam G, Zeevi D, Qureshi A, Malhotra A, Majidi S, Murthy S, Park S, Kontos D, Falcone G, Sheth K, Payabvash S. Improving the Robustness of Deep Learning Models in Predicting Hematoma Expansion from Admission Head CT. American Journal Of Neuroradiology 2025, 46: 1404-1411. PMID: 39794133, PMCID: PMC12453420, DOI: 10.3174/ajnr.a8650.Peer-Reviewed Original ResearchFast Gradient Sign MethodDeep learning modelsRobustness of deep learning modelsAdversarial attacksAdversarial imagesAdversarial trainingSign methodModel robustnessDeploying deep learning modelsDeep learning model performanceConvolutional neural networkImprove model robustnessAcute intracerebral hemorrhageHematoma expansionMulti-threshold segmentationReceiver Operating CharacteristicIntracerebral hemorrhageGradient descentType attacksData perturbationNeural networkProjected GradientTraining setAntihypertensive Treatment of Acute Cerebral HemorrhageThreshold-based segmentation
This site is protected by hCaptcha and its Privacy Policy and Terms of Service apply