Featured Publications
Unsupervised inter-frame motion correction for whole-body dynamic PET using convolutional long short-term memory in a convolutional neural network
Guo X, Zhou B, Pigg D, Spottiswoode B, Casey ME, Liu C, Dvornek NC. Unsupervised inter-frame motion correction for whole-body dynamic PET using convolutional long short-term memory in a convolutional neural network. Medical Image Analysis 2022, 80: 102524. PMID: 35797734, PMCID: PMC10923189, DOI: 10.1016/j.media.2022.102524.Peer-Reviewed Original ResearchConceptsConvolutional neural networkNeural networkConvolutional long short-term memory (ConvLSTM) layersDeep learning-based frameworkConvolutional long short-term memoryLong short-term memory layersDeep learning baselinesLong short-term memoryDynamic temporal featuresLearning-based frameworkDeep learning approachShort-term memory layersTracer distribution changeMotion estimation networkMotion prediction errorInference timeEstimation networkLearning baselinesNon-rigid registration methodLearning approachMotion correction methodMemory layerShort-term memoryTemporal featuresRegistration method
2024
TAI-GAN: A Temporally and Anatomically Informed Generative Adversarial Network for early-to-late frame conversion in dynamic cardiac PET inter-frame motion correction
Guo X, Shi L, Chen X, Liu Q, Zhou B, Xie H, Liu Y, Palyo R, Miller E, Sinusas A, Staib L, Spottiswoode B, Liu C, Dvornek N. TAI-GAN: A Temporally and Anatomically Informed Generative Adversarial Network for early-to-late frame conversion in dynamic cardiac PET inter-frame motion correction. Medical Image Analysis 2024, 96: 103190. PMID: 38820677, PMCID: PMC11180595, DOI: 10.1016/j.media.2024.103190.Peer-Reviewed Original ResearchGenerative adversarial networkAdversarial networkMotion estimation accuracyInter-frame motionIntensity-based image registration techniqueAll-to-oneSegmentation masksImage registration techniquesOriginal frameTemporal informationDiagnosis accuracyMyocardial blood flowEstimation accuracyFrame conversionPositron emission tomographyNovel methodImage qualityPET datasetsRegistration techniqueNetworkCardiac positron emission tomographyBlood flowDynamic cardiac positron emission tomographyMotion correctionCoronary artery diseaseChapter 13 Data-driven learning strategies for biomarker detection and outcome prediction in Autism from task-based fMRI
Duncan J, Staib L, Dvornek N, Li X, Zhuang J, Wang J, Ventola P. Chapter 13 Data-driven learning strategies for biomarker detection and outcome prediction in Autism from task-based fMRI. 2024, 357-393. DOI: 10.1016/b978-0-32-385124-4.00024-6.Peer-Reviewed Original ResearchLong short-term memoryGraph neural networksFunctional magnetic resonance imagingAutism spectrum disorderNeural ordinary differential equationsData-driven learning strategyDeep learning techniquesTask-based functional magnetic resonance imagingShort-term memoryNeural networkLearning techniquesImpaired social interactionTerm memoryBehavioral therapyRepetitive behaviorsSpectrum disorderDevelopmental disordersLearning strategiesSpatio-temporal characteristicsInherent dynamicsCharacterization of individualsModel of causalitySocial interactionNetworkPersonalized outcome predictions
2023
TAI-GAN: Temporally and Anatomically Informed GAN for Early-to-Late Frame Conversion in Dynamic Cardiac PET Motion Correction
Guo X, Shi L, Chen X, Zhou B, Liu Q, Xie H, Liu Y, Palyo R, Miller E, Sinusas A, Spottiswoode B, Liu C, Dvornek N. TAI-GAN: Temporally and Anatomically Informed GAN for Early-to-Late Frame Conversion in Dynamic Cardiac PET Motion Correction. Lecture Notes In Computer Science 2023, 14288: 64-74. PMID: 38464964, PMCID: PMC10923183, DOI: 10.1007/978-3-031-44689-4_7.Peer-Reviewed Original ResearchLearning Sequential Information in Task-Based fMRI for Synthetic Data Augmentation
Wang J, Dvornek N, Staib L, Duncan J. Learning Sequential Information in Task-Based fMRI for Synthetic Data Augmentation. Lecture Notes In Computer Science 2023, 14312: 79-88. DOI: 10.1007/978-3-031-44858-4_8.Peer-Reviewed Original ResearchFunctional magnetic resonance imagesData augmentationClassification taskSpecific cognitive tasksMedical image analysisSynthetic data augmentationEffective data augmentationDownstream learning tasksCognitive tasksVariational autoencoder modelLearning taskTraining dataAutoencoder modelTemporal informationTraining datasetSequential informationSynthetic imagesTaskFMRI sequencesImage analysisMultiple perspectivesMagnetic resonance imagesImagesDifferent alternativesPersistent issueCopy Number Variation Informs fMRI-Based Prediction of Autism Spectrum Disorder
Dvornek N, Sullivan C, Duncan J, Gupta A. Copy Number Variation Informs fMRI-Based Prediction of Autism Spectrum Disorder. Lecture Notes In Computer Science 2023, 14312: 133-142. PMID: 38371906, PMCID: PMC10868600, DOI: 10.1007/978-3-031-44858-4_13.Peer-Reviewed Original ResearchMCP-Net: Introducing Patlak Loss Optimization to Whole-Body Dynamic PET Inter-Frame Motion Correction
Guo X, Zhou B, Chen X, Chen M, Liu C, Dvornek N. MCP-Net: Introducing Patlak Loss Optimization to Whole-Body Dynamic PET Inter-Frame Motion Correction. IEEE Transactions On Medical Imaging 2023, 42: 3512-3523. PMID: 37368811, PMCID: PMC10751388, DOI: 10.1109/tmi.2023.3290003.Peer-Reviewed Original ResearchMotion estimation blockDeep learning benchmarksGood generalization capabilityMotion correctionMotion correction frameworkMotion prediction errorGeneralization capabilityNetwork performanceNeural networkMotion correction techniqueLearning benchmarksRegistration problemLoss functionEstimation blockLoss optimizationPenalty componentDynamic frameFitting errorSpatial alignmentParametric imagesSpatial misalignmentDynamic positron emission tomographySubject motionPrediction errorCorrection frameworkChapter 13 Deep learning with connectomes
Dvornek N, Li X. Chapter 13 Deep learning with connectomes. 2023, 289-308. DOI: 10.1016/b978-0-323-85280-7.00013-0.ChaptersDeep learning modelsLearning modelDeep learningClassic computer visionNeural network architectureImage analysis problemsMachine learning methodsNeural network modelComputer visionPotential future workNetwork architectureNonlinear neural network modelArt resultsPrediction taskLearning methodsNetwork modelAnalysis problemUseful representationConnectomePopular typeLearningFuture workData analysisArchitectureTask
2022
Inter-Pass Motion Correction for Whole-Body Dynamic PET and Parametric Imaging
Guo X, Wu J, Chen M, Liu Q, Onofrey J, Pucar D, Pang Y, Pigg D, Casey M, Dvornek N, Liu C. Inter-Pass Motion Correction for Whole-Body Dynamic PET and Parametric Imaging. IEEE Transactions On Radiation And Plasma Medical Sciences 2022, 7: 344-353. PMID: 37842204, PMCID: PMC10569406, DOI: 10.1109/trpms.2022.3227576.Peer-Reviewed Original ResearchMCP-Net: Inter-frame Motion Correction with Patlak Regularization for Whole-body Dynamic PET
Guo X, Zhou B, Chen X, Liu C, Dvornek N. MCP-Net: Inter-frame Motion Correction with Patlak Regularization for Whole-body Dynamic PET. Lecture Notes In Computer Science 2022, 13434: 163-172. PMID: 38464686, PMCID: PMC10923180, DOI: 10.1007/978-3-031-16440-8_16.Peer-Reviewed Original ResearchConvolutional long short-term memory (ConvLSTM) layersLong short-term memory layersMotion estimation moduleShort-term memory layersDeep learning benchmarksEnhanced network performanceImage registration problemMotion correction frameworkMotion correctionU-NetNetwork performanceLearning benchmarksSimilarity measurementEstimation moduleRegistration problemGradient lossMemory layerLoss functionDynamic frameDynamic positron emission tomographyFitting errorSpatial alignmentSpatial misalignmentPatient motionModuleCharacterization of Early Stage Parkinson's Disease From Resting-State fMRI Data Using a Long Short-Term Memory Network
Guo X, Tinaz S, Dvornek N. Characterization of Early Stage Parkinson's Disease From Resting-State fMRI Data Using a Long Short-Term Memory Network. Frontiers In Neuroimaging 2022, 1: 952084. PMID: 37555151, PMCID: PMC10406199, DOI: 10.3389/fnimg.2022.952084.Peer-Reviewed Original ResearchEarly-stage Parkinson's diseaseFunctional magnetic resonance imagingParkinson's Progression Markers InitiativeParkinson's diseaseProgression Markers InitiativeDiagnosis of PDEarly-stage diseaseFunctional brain changesBrain function alterationsStage Parkinson's diseaseFunctional connectivity differencesComplex neurodegenerative disorderMagnetic resonance imagingResting-state fMRI dataStage diseaseDisease stageDisease progressionBrain changesTreatment responseMotor impairmentFC changesNew therapiesFunction alterationsResonance imagingBrain regions
2021
Automatic Inter-Frame Patient Motion Correction for Dynamic Cardiac PET Using Deep Learning
Shi L, Lu Y, Dvornek N, Weyman CA, Miller EJ, Sinusas AJ, Liu C. Automatic Inter-Frame Patient Motion Correction for Dynamic Cardiac PET Using Deep Learning. IEEE Transactions On Medical Imaging 2021, 40: 3293-3304. PMID: 34018932, PMCID: PMC8670362, DOI: 10.1109/tmi.2021.3082578.Peer-Reviewed Original ResearchConceptsConvolutional neural networkRegistration-based methodMotion correctionDynamic frameTracer distribution changeDynamic image dataPatient motion correctionPatient scansDeep learningPatient motionMotion estimationImage dataLSTM networkNeural networkRealistic patient motionTemporal informationMotion correction methodMotion detectionCardiac PETClinical workflowRigid translational motionFlow estimationNetworkPatient datasetsSuperior performanceBrainGNN: Interpretable Brain Graph Neural Network for fMRI Analysis
Li X, Zhou Y, Dvornek N, Zhang M, Gao S, Zhuang J, Scheinost D, Staib LH, Ventola P, Duncan JS. BrainGNN: Interpretable Brain Graph Neural Network for fMRI Analysis. Medical Image Analysis 2021, 74: 102233. PMID: 34655865, PMCID: PMC9916535, DOI: 10.1016/j.media.2021.102233.Peer-Reviewed Original ResearchConceptsFunctional magnetic resonance imagesGraph neural network frameworkMedical image analysisGraph neural networkGraph convolutional layersNeural network frameworkDifferent evaluation metricsSpecific task statesIndependent fMRI datasetsPooling layerConvolutional layersConsistency lossNetwork frameworkNeural networkFMRI datasetsImage analysis methodEvaluation metricsDetection resultsBrain graphsSubjects releaseROI selectionImage analysisCognitive stimuliTask statesFMRI analysisMultiple-Shooting Adjoint Method for Whole-Brain Dynamic Causal Modeling
Zhuang J, Dvornek N, Tatikonda S, Papademetris X, Ventola P, Duncan J. Multiple-Shooting Adjoint Method for Whole-Brain Dynamic Causal Modeling. Lecture Notes In Computer Science 2021, 12729: 58-70. DOI: 10.1007/978-3-030-78191-0_5.Peer-Reviewed Original ResearchOrdinary differential equationsAdjoint methodNoisy observationsMultiple shooting methodNon-linear systemsLarge scale continuous systemsLarge-scale systemsParameter value estimationDifferential equationsAccurate gradient estimationExpectation-maximization algorithmNon-linear modelParameter estimationBayesian frameworkGradient estimationContinuous systemToy exampleLarge systemsReal fMRI dataEstimationValue estimationAlgorithmGood accuracyCausal modelingModel changesA Metamodel Structure For Regression Analysis: Application To Prediction Of Autism Spectrum Disorder Severity
Wang S, Dvornek N. A Metamodel Structure For Regression Analysis: Application To Prediction Of Autism Spectrum Disorder Severity. 2021, 00: 1338-1341. DOI: 10.1109/isbi48211.2021.9434009.Peer-Reviewed Original ResearchNeuropsychiatric disease classification using functional connectomics - results of the connectomics in neuroimaging transfer learning challenge
Schirmer MD, Venkataraman A, Rekik I, Kim M, Mostofsky SH, Nebel MB, Rosch K, Seymour K, Crocetti D, Irzan H, Hütel M, Ourselin S, Marlow N, Melbourne A, Levchenko E, Zhou S, Kunda M, Lu H, Dvornek NC, Zhuang J, Pinto G, Samal S, Zhang J, Bernal-Rusiel JL, Pienaar R, Chung AW. Neuropsychiatric disease classification using functional connectomics - results of the connectomics in neuroimaging transfer learning challenge. Medical Image Analysis 2021, 70: 101972. PMID: 33677261, PMCID: PMC9115580, DOI: 10.1016/j.media.2021.101972.Peer-Reviewed Original ResearchConceptsAttention-deficit/hyperactivity disorderAutism Brain Imaging Data ExchangeResting-state fMRI time seriesAutism spectrum disorder (ASD) patientsHuman Connectome ProjectADHD modelHyperactivity disorderADHD comorbidityASD classificationConnectome ProjectOpen-source image analysis platformFMRI time seriesBrain connectomicsFunctional connectomicsClassification methodologyLearning challengesOpen-source datasetField of connectomicsParcellation atlasesDisorder patientsImage analysis platformOmission rateParticipantsConnectomicsMICCAI 2019
2020
Cross-Modality Segmentation by Self-supervised Semantic Alignment in Disentangled Content Space
Yang J, Li X, Pak D, Dvornek N, Chapiro J, Lin M, Duncan J. Cross-Modality Segmentation by Self-supervised Semantic Alignment in Disentangled Content Space. Lecture Notes In Computer Science 2020, 12444: 52-61. DOI: 10.1007/978-3-030-60548-3_6.Peer-Reviewed Original ResearchSemantic alignmentDomain shiftContent spaceCross-modality medical image segmentationUnsupervised domain adaptation problemMedical image tasksMedical image segmentationDeep convolutional networksDomain adaptation problemAdversarial training strategyAdvantage of imagesQualitative experimental resultsDifferent modalitiesDeep networkPretext taskConvolutional networkSegmentation taskImage segmentationImage tasksSegmentation performanceArt performanceLiver segmentationModality alignmentGeneralization abilitySegmentation pipeline
2019
Deep Learning based Respiratory Pattern Classification and Applications in PET/CT Motion Correction
Guo Y, Dvornek N, Lu Y, Tsai Y, Hamill J, Casey M, Liu C. Deep Learning based Respiratory Pattern Classification and Applications in PET/CT Motion Correction. 2019, 00: 1-5. DOI: 10.1109/nss/mic42101.2019.9059783.Peer-Reviewed Original ResearchDeep learningNeural networkMotion correction methodDeep neural networksDeep learning modelsHybrid neural networkConvolutional layersHigh prediction accuracyRecurrent layersGeneralization capabilityData preprocessingLearning modelPattern classificationRespiratory motionAnzai systemLoss functionLinear classifierPrediction accuracyIntra-gate motionRPM systemMotion correctionTumor detectionNetworkIrregular breathersCT imagesShelfNet for Fast Semantic Segmentation
Zhuang J, Yang J, Gu L, Dvornek N. ShelfNet for Fast Semantic Segmentation. 2019, 00: 847-856. DOI: 10.1109/iccvw.2019.00113.Peer-Reviewed Original ResearchFast semantic segmentationSemantic segmentationCityscapes datasetReal-time segmentation modelStreet-scene understandingFast inference speedEncoder-decoder structurePASCAL VOC datasetHigh accuracyInference speedSkip connectionsAutonomous drivingExtensive experimentsResidual blocksSegmentation modelVOC datasetNovel architectureReal-time methodComputation burdenShallow pathsBiSeNetPSPNetParameter numberComparable speedDataset
2018
2-Channel Convolutional 3D Deep Neural Network (2CC3D) for FMRI Analysis: ASD Classification and Feature Learning
Li X, Dvornek NC, Papademetris X, Zhuang J, Staib LH, Ventola P, Duncan JS. 2-Channel Convolutional 3D Deep Neural Network (2CC3D) for FMRI Analysis: ASD Classification and Feature Learning. 2011 IEEE International Symposium On Biomedical Imaging: From Nano To Macro 2018, 2018: 1252-1255. PMID: 32983370, PMCID: PMC7519578, DOI: 10.1109/isbi.2018.8363798.Peer-Reviewed Original ResearchConvolutional neural networkNeural networkCNN convolutional layerSpatial featuresASD classificationDeep neural networksMean F-scoreTraditional machineFeature learningConvolutional layersInput formatF-scoreClassification modelTemporal informationNetworkWindow parametersImagesClassificationConvolutionalTemporal statisticsMachineLearningFeaturesFormatScheme