Andreas Coppi
Associate Research Scientist (Cardiovascular Medicine)Cards
About
Research
Publications
2026
Artificial intelligence-based automated interpretation of images of electrocardiograms: development and multinational validation of ECG-GPT
Khunte A, Sangha V, Oikonomou E, Dhingra L, Aminorroaya A, Coppi A, Shankar S, Rockers E, Mortazavi B, Bhatt D, Krumholz H, Al-Kindi S, Nadkarni G, Vaid A, Khera R. Artificial intelligence-based automated interpretation of images of electrocardiograms: development and multinational validation of ECG-GPT. European Heart Journal - Digital Health 2026, 7: ztag031. PMID: 41853639, PMCID: PMC12993923, DOI: 10.1093/ehjdh/ztag031.Peer-Reviewed Original ResearchBundle branch blockEncoder-decoder modelBranch blockFascicular blockUS health systemClinical assessmentLeft anterior fascicular blockRight bundle branch blockLeft posterior fascicular blockAssessment of electrocardiogramsLeft bundle branch blockAnterior fascicular blockPosterior fascicular blockPremature atrial contractionsPremature ventricular contractionsHealth systemPTB-XL datasetLow-resource settingsAtrioventricular blockConduction abnormalitiesSinus bradycardiaAtrial contractionSinus tachycardiaAtrial fibrillationDiagnosis statementsTARGET-AI: A Foundational Approach for the Targeted Deployment of Artificial Intelligence Electrocardiography in the Electronic Health Record
Oikonomou E, Batinica B, Dhingra L, Aminorroaya A, Coppi A, Khera R. TARGET-AI: A Foundational Approach for the Targeted Deployment of Artificial Intelligence Electrocardiography in the Electronic Health Record. NEJM AI 2026, 3 DOI: 10.1056/aioa2500588.Peer-Reviewed Original ResearchClinical Trial Patient Matching: A Real-Time, Common Data Model and Artificial Intelligence–Driven System for Semiautomated Patient Prescreening in Cancer Clinical Trials
Gong G, Liu J, Pandya S, Taborda C, Wiesendanger N, Price N, Byron W, Coppi A, Young P, Wiess C, Dunning H, Barganier C, Brodeur R, Fischbach N, LoRusso P, Pusztai L, Kim S, Rozenblit M, Cecchini M, Mongiu A, Mendez L, Kaftan E, Torre C, Krumholz H, Krop I, Schulz W, Lustberg M, Kunz P. Clinical Trial Patient Matching: A Real-Time, Common Data Model and Artificial Intelligence–Driven System for Semiautomated Patient Prescreening in Cancer Clinical Trials. JCO Clinical Cancer Informatics 2026, 10: e2500262. PMID: 41512229, DOI: 10.1200/cci-25-00262.Peer-Reviewed Original ResearchConceptsObservational Medical Outcomes PartnershipHealth systemColorectal cancerElectronic health record dataCancer clinical trial enrollmentChart reviewHealth record dataManual chart reviewClinical trial recruitmentClinical trialsCancer clinical trialsCancer specialtiesCommon data modelScreen timeColorectal cancer trialsClinical trial enrollmentTrial recruitmentClinical chart reviewConsent ratesPatient accessExhaustive chart reviewMetastatic colorectal cancerEnrollment challengesRecord dataTrial enrollmentAssessment of the integrity of real-time electronic health record data used in clinical research
Liu J, Pandya S, Coppi A, Young H, Krumholz H, Schulz W, Gong G. Assessment of the integrity of real-time electronic health record data used in clinical research. PLOS ONE 2026, 21: e0340287. PMID: 41511976, PMCID: PMC12788664, DOI: 10.1371/journal.pone.0340287.Peer-Reviewed Original ResearchConceptsElectronic health recordsEHR dataReal-time electronic health recordsElectronic health record dataSecondary useHealth record dataClinical actionsIntegration of real-time dataClinical trial readinessCommon data modelHealth recordsHealth systemOMOP Common Data ModelDischarge informationClinical careResearch readinessRecord dataTrial readinessSynthetic datasetsEHR datasetData warehouseDemographic variablesReal-time dataPost-encounterAutomated framework
2025
Targeted deployment of AI-ECG for efficient screening of transthyretin amyloid cardiomyopathy using deep learning representations of longitudinal electronic health records
Oikonomou E, Dhingra L, Batinica B, Coppi A, Malicki C, Pedroso A, Khera R. Targeted deployment of AI-ECG for efficient screening of transthyretin amyloid cardiomyopathy using deep learning representations of longitudinal electronic health records. European Heart Journal 2025, 46 DOI: 10.1093/eurheartj/ehaf784.2685.Peer-Reviewed Original ResearchElectronic health recordsDeep learning representationsLongitudinal electronic health recordsHealth systemHealth recordsIndividual electronic health recordsOptimal decision thresholdIndividuals seeking careOpportunistic deploymentTargeted deploymentDeep learningHigh precisionOptimal deploymentTraining setDownstream testingHealthcare encountersMultimodal pipelinePositive screenDecision thresholdDeploymentATTR-CMOptimal intersectionDevelopment setHealthRepresentationGuiding the targeted deployment of AI-ECG for the precision diagnosis of structural heart disorders in the electronic health record
Oikonomou E, Batinica B, Dhingra L, Aminorroaya A, Coppi A, Khera R. Guiding the targeted deployment of AI-ECG for the precision diagnosis of structural heart disorders in the electronic health record. European Heart Journal 2025, 46 DOI: 10.1093/eurheartj/ehaf784.4470.Peer-Reviewed Original ResearchElectronic health recordsHealth recordsF1 scoreElectrocardiogram imageEHR eventContrastive Language-Image Pre-trainingTest setDeep learning representationsFalse positive screensLongitudinal EHR dataStructural heart disordersBalance of precisionLearned representationsTrained embeddingsImage encoderVision TransformerUK BiobankHealth systemEHR dataPositive screenAlgorithmic pipelinePre-trainingAI-ECGDeploymentHeart diseaseArtificial intelligence-enabled electrocardiography and echocardiography to track preclinical progression of transthyretin amyloid cardiomyopathy
Oikonomou E, Sangha V, Shankar S, Coppi A, Krumholz H, Nasir K, Miller E, Gallegos Kattan C, Al-Mallah M, Al-Kindi S, Khera R. Artificial intelligence-enabled electrocardiography and echocardiography to track preclinical progression of transthyretin amyloid cardiomyopathy. European Heart Journal 2025, 46: 3651-3662. PMID: 40679604, PMCID: PMC12488324, DOI: 10.1093/eurheartj/ehaf450.Peer-Reviewed Original ResearchTransthyretin amyloid cardiomyopathyTransthoracic echocardiographyATTR-CMAmyloid cardiomyopathyPreclinical progressAI-ECGRetrospective analysisDiagnosis of transthyretin amyloid cardiomyopathyDeep learning modelsAge/sex matched controlsRetrospective analysis of individualsLearning modelsPreclinical testingElectrocardiography imagingEchocardiographyHouston Methodist HospitalYale New Haven Health SystemAdvanced imagingElectrocardiographyPreclinical coursesCardiomyopathyPreclinical stageDevelopment and multinational validation of an ensemble deep learning algorithm for detecting and predicting structural heart disease using noisy single-lead electrocardiograms
Aminorroaya A, Dhingra L, Pedroso A, Shankar S, Coppi A, Khunte A, Foppa M, Brant L, Barreto S, Ribeiro A, Krumholz H, Oikonomou E, Khera R. Development and multinational validation of an ensemble deep learning algorithm for detecting and predicting structural heart disease using noisy single-lead electrocardiograms. European Heart Journal - Digital Health 2025, 6: 554-566. PMID: 40703117, PMCID: PMC12282373, DOI: 10.1093/ehjdh/ztaf034.Peer-Reviewed Original ResearchDetectable structural heart diseaseStructural heart diseaseCommunity-based screeningLeft-sided valvular diseaseHeart diseaseELSA-BrasilYale-New Haven HospitalAI-ECG algorithmDeep learning algorithmsPopulation-based cohortSevere LVHEchocardiographic dataPredictive biomarkersHospital-based sitesNew Haven HospitalRisk stratificationValvular diseaseEnsemble deep learning algorithmUK BiobankCommunity hospitalLead I ECGNirmatrelvir–ritonavir versus placebo–ritonavir in individuals with long COVID in the USA (PAX LC): a double-blind, randomised, placebo-controlled, phase 2, decentralised trial
Sawano M, Bhattacharjee B, Caraballo C, Khera R, Li S, Herrin J, Christian D, Coppi A, Warner F, Holub J, Henriquez Y, Johnson M, Goddard T, Rocco E, Hummel A, Mouslmani M, Hooper W, Putrino D, Carr K, Charnas L, De Jesus M, Nepert D, Abreu P, Ziegler F, Spertus J, Iwasaki A, Krumholz H. Nirmatrelvir–ritonavir versus placebo–ritonavir in individuals with long COVID in the USA (PAX LC): a double-blind, randomised, placebo-controlled, phase 2, decentralised trial. The Lancet Infectious Diseases 2025, 25: 936-946. PMID: 40188838, DOI: 10.1016/s1473-3099(25)00073-8.Peer-Reviewed Original ResearchPhysical health summary scoreBaseline to dayAdverse eventsNirmatrelvir-ritonavirSARS-CoV-2 infectionDouble-blindStudy drug-related treatment-emergent adverse eventsDrug-related treatment-emergent adverse eventsTreatment-emergent adverse eventsIntention-to-treat populationWeek 6Baseline to week 6Documented SARS-CoV-2 infectionActive liver diseaseEffective pharmacological interventionsLong COVIDAcute medical illnessSafety populationPatient-Reported Outcomes Measurement Information SystemEarly treatment terminationRenal impairmentTreat long-COVIDPlacebo-ControlledEfficacy endpointRandomised controlled trialsEnsemble Deep Learning Algorithm for Structural Heart Disease Screening Using Electrocardiographic Images PRESENT SHD
Dhingra L, Aminorroaya A, Sangha V, Pedroso A, Shankar S, Coppi A, Foppa M, Brant L, Barreto S, Ribeiro A, Krumholz H, Oikonomou E, Khera R. Ensemble Deep Learning Algorithm for Structural Heart Disease Screening Using Electrocardiographic Images PRESENT SHD. Journal Of The American College Of Cardiology 2025, 85: 1302-1313. PMID: 40139886, PMCID: PMC12199746, DOI: 10.1016/j.jacc.2025.01.030.Peer-Reviewed Original ResearchConceptsStructural heart diseaseYale-New Haven HospitalTransthoracic echocardiogramRisk stratificationHeart failureLeft-sided valvular diseaseSevere left ventricular hypertrophyLeft ventricular ejection fractionReceiver-operating characteristic curveVentricular ejection fractionLeft ventricular hypertrophyHeart disease screeningELSA-BrasilEnsemble deep learning algorithmRisk of deathConvolutional neural network modelEjection fractionEnsemble deep learning approachVentricular hypertrophyDeep learning algorithmsNew Haven HospitalDeep learning approachValvular diseaseNeural network modelClinical cohort