Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in context
Summary: Background: Atherosclerotic cardiovascular disease (ASCVD) is the leading cause of death worldwide, driven primarily by coronary artery disease (CAD). ASCVD risk estimators such as the pooled cohort equations (PCE) facilitate risk stratification and primary prevention of ASCVD but their ac...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-11-01
|
| Series: | EClinicalMedicine |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2589537023004364 |
| _version_ | 1827631719072661504 |
|---|---|
| author | Samir Awasthi Nikhil Sachdeva Yash Gupta Ausath G. Anto Shahir Asfahan Ruben Abbou Sairam Bade Sanyam Sood Lars Hegstrom Nirupama Vellanki Heather M. Alger Melwin Babu Jose R. Medina-Inojosa Robert B. McCully Amir Lerman Mark Stampehl Rakesh Barve Zachi I. Attia Paul A. Friedman Venky Soundararajan Francisco Lopez-Jimenez |
| author_facet | Samir Awasthi Nikhil Sachdeva Yash Gupta Ausath G. Anto Shahir Asfahan Ruben Abbou Sairam Bade Sanyam Sood Lars Hegstrom Nirupama Vellanki Heather M. Alger Melwin Babu Jose R. Medina-Inojosa Robert B. McCully Amir Lerman Mark Stampehl Rakesh Barve Zachi I. Attia Paul A. Friedman Venky Soundararajan Francisco Lopez-Jimenez |
| author_sort | Samir Awasthi |
| collection | DOAJ |
| description | Summary: Background: Atherosclerotic cardiovascular disease (ASCVD) is the leading cause of death worldwide, driven primarily by coronary artery disease (CAD). ASCVD risk estimators such as the pooled cohort equations (PCE) facilitate risk stratification and primary prevention of ASCVD but their accuracy is still suboptimal. Methods: Using deep electronic health record data from 7,116,209 patients seen at 70+ hospitals and clinics across 5 states in the USA, we developed an artificial intelligence-based electrocardiogram analysis tool (ECG-AI) to detect CAD and assessed the additive value of ECG-AI-based ASCVD risk stratification to the PCE. We created independent ECG-AI models using separate neural networks including subjects without known history of ASCVD, to identify coronary artery calcium (CAC) score ≥300 Agatston units by computed tomography, obstructive CAD by angiography or procedural intervention, and regional left ventricular akinesis in ≥1 segment by echocardiogram, as a reflection of possible prior myocardial infarction (MI). These were used to assess the utility of ECG-AI-based ASCVD risk stratification in a retrospective observational study consisting of patients with PCE scores and no prior ASCVD. The study period covered all available digitized EHR data, with the first available ECG in 1987 and the last in February 2023. Findings: ECG-AI for identifying CAC ≥300, obstructive CAD, and regional akinesis achieved area under the receiver operating characteristic (AUROC) values of 0.88, 0.85, and 0.94, respectively. An ensembled ECG-AI identified 3, 5, and 10-year risk for acute coronary events and mortality independently and additively to PCE. Hazard ratios for acute coronary events over 3-years in patients without ASCVD that tested positive on 1, 2, or 3 versus 0 disease-specific ECG-AI models at cohort entry were 2.41 (2.14–2.71), 4.23 (3.74–4.78), and 11.75 (10.2–13.52), respectively. Similar stratification was observed in cohorts stratified by PCE or age. Interpretation: ECG-AI has potential to address unmet need for accessible risk stratification in patients in whom PCE under, over, or insufficiently estimates ASCVD risk, and in whom risk assessment over time periods shorter than 10 years is desired. Funding: Anumana. |
| first_indexed | 2024-03-09T14:25:58Z |
| format | Article |
| id | doaj.art-97d6bc637a9f484f91ae2a92cbf0ea4a |
| institution | Directory Open Access Journal |
| issn | 2589-5370 |
| language | English |
| last_indexed | 2024-03-09T14:25:58Z |
| publishDate | 2023-11-01 |
| publisher | Elsevier |
| record_format | Article |
| series | EClinicalMedicine |
| spelling | doaj.art-97d6bc637a9f484f91ae2a92cbf0ea4a2023-11-28T07:26:43ZengElsevierEClinicalMedicine2589-53702023-11-0165102259Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in contextSamir Awasthi0Nikhil Sachdeva1Yash Gupta2Ausath G. Anto3Shahir Asfahan4Ruben Abbou5Sairam Bade6Sanyam Sood7Lars Hegstrom8Nirupama Vellanki9Heather M. Alger10Melwin Babu11Jose R. Medina-Inojosa12Robert B. McCully13Amir Lerman14Mark Stampehl15Rakesh Barve16Zachi I. Attia17Paul A. Friedman18Venky Soundararajan19Francisco Lopez-Jimenez20Anumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAnference, Inc, One Main Street, Cambridge, MA, USA; Beth Israel Deaconess Medical Center, Boston, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAMayo Clinic, Rochester, MN, USAMayo Clinic, Rochester, MN, USAMayo Clinic, Rochester, MN, USANovartis Pharmaceuticals Corporation, East Hanover, NJ, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAMayo Clinic, Rochester, MN, USAMayo Clinic, Rochester, MN, USAAnumana, Inc, One Main Street, Cambridge, MA, USA; nference, Inc, One Main Street, Cambridge, MA, USAMayo Clinic, Rochester, MN, USA; Corresponding author.Summary: Background: Atherosclerotic cardiovascular disease (ASCVD) is the leading cause of death worldwide, driven primarily by coronary artery disease (CAD). ASCVD risk estimators such as the pooled cohort equations (PCE) facilitate risk stratification and primary prevention of ASCVD but their accuracy is still suboptimal. Methods: Using deep electronic health record data from 7,116,209 patients seen at 70+ hospitals and clinics across 5 states in the USA, we developed an artificial intelligence-based electrocardiogram analysis tool (ECG-AI) to detect CAD and assessed the additive value of ECG-AI-based ASCVD risk stratification to the PCE. We created independent ECG-AI models using separate neural networks including subjects without known history of ASCVD, to identify coronary artery calcium (CAC) score ≥300 Agatston units by computed tomography, obstructive CAD by angiography or procedural intervention, and regional left ventricular akinesis in ≥1 segment by echocardiogram, as a reflection of possible prior myocardial infarction (MI). These were used to assess the utility of ECG-AI-based ASCVD risk stratification in a retrospective observational study consisting of patients with PCE scores and no prior ASCVD. The study period covered all available digitized EHR data, with the first available ECG in 1987 and the last in February 2023. Findings: ECG-AI for identifying CAC ≥300, obstructive CAD, and regional akinesis achieved area under the receiver operating characteristic (AUROC) values of 0.88, 0.85, and 0.94, respectively. An ensembled ECG-AI identified 3, 5, and 10-year risk for acute coronary events and mortality independently and additively to PCE. Hazard ratios for acute coronary events over 3-years in patients without ASCVD that tested positive on 1, 2, or 3 versus 0 disease-specific ECG-AI models at cohort entry were 2.41 (2.14–2.71), 4.23 (3.74–4.78), and 11.75 (10.2–13.52), respectively. Similar stratification was observed in cohorts stratified by PCE or age. Interpretation: ECG-AI has potential to address unmet need for accessible risk stratification in patients in whom PCE under, over, or insufficiently estimates ASCVD risk, and in whom risk assessment over time periods shorter than 10 years is desired. Funding: Anumana.http://www.sciencedirect.com/science/article/pii/S2589537023004364Artificial intelligenceECG-AICoronary artery diseaseAtherosclerotic cardiovascular diseaseCardiovascular risk |
| spellingShingle | Samir Awasthi Nikhil Sachdeva Yash Gupta Ausath G. Anto Shahir Asfahan Ruben Abbou Sairam Bade Sanyam Sood Lars Hegstrom Nirupama Vellanki Heather M. Alger Melwin Babu Jose R. Medina-Inojosa Robert B. McCully Amir Lerman Mark Stampehl Rakesh Barve Zachi I. Attia Paul A. Friedman Venky Soundararajan Francisco Lopez-Jimenez Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in context EClinicalMedicine Artificial intelligence ECG-AI Coronary artery disease Atherosclerotic cardiovascular disease Cardiovascular risk |
| title | Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in context |
| title_full | Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in context |
| title_fullStr | Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in context |
| title_full_unstemmed | Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in context |
| title_short | Identification and risk stratification of coronary disease by artificial intelligence-enabled ECGResearch in context |
| title_sort | identification and risk stratification of coronary disease by artificial intelligence enabled ecgresearch in context |
| topic | Artificial intelligence ECG-AI Coronary artery disease Atherosclerotic cardiovascular disease Cardiovascular risk |
| url | http://www.sciencedirect.com/science/article/pii/S2589537023004364 |
| work_keys_str_mv | AT samirawasthi identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT nikhilsachdeva identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT yashgupta identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT ausathganto identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT shahirasfahan identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT rubenabbou identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT sairambade identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT sanyamsood identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT larshegstrom identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT nirupamavellanki identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT heathermalger identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT melwinbabu identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT josermedinainojosa identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT robertbmccully identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT amirlerman identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT markstampehl identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT rakeshbarve identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT zachiiattia identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT paulafriedman identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT venkysoundararajan identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext AT franciscolopezjimenez identificationandriskstratificationofcoronarydiseasebyartificialintelligenceenabledecgresearchincontext |