{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,2]],"date-time":"2026-05-02T15:38:14Z","timestamp":1777736294561,"version":"3.51.4"},"reference-count":49,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,3,17]],"date-time":"2025-03-17T00:00:00Z","timestamp":1742169600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec><jats:title>Background<\/jats:title><jats:p>The diagnostic power of exercise stress electrocardiography (ExECG) remains limited. We aimed to construct an artificial intelligence (AI)-based method to enhance ExECG performance to identify patients with significant coronary artery disease (CAD).<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>We retrospectively collected 818 patients who underwent both ExECG and coronary angiography (CAG) within 6\u202fmonths. The mean age was 57.0\u202f\u00b1\u202f10.1\u202fyears, and 614 (75%) were male patients. Significant coronary artery disease was seen in 369 (43.8%) CAG reports. We also included 197 individuals with normal ExECG and low risk of CAD. A convolutional recurrent neural network algorithm, integrating electrocardiographic (ECG) signals and features from ExECG reports, was developed to predict the risk of significant CAD. We also investigated the optimal number of inputted ECG signal slices and features and the weighting of features for model performance.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Using the data of patients undergoing CAG for training and test sets, our algorithm had an area under the curve, sensitivity, and specificity of 0.74, 0.86, and 0.47, respectively, which increased to 0.83, 0.89, and 0.60, respectively, after enrolling 197 subjects with low risk of CAD. Three ECG signal slices and 12 features yielded optimal performance metrics. The principal predictive feature variables were sex, maximum heart rate, and ST\/HR index. Our model generated results within one minute after completing ExECG.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>The multimodal AI algorithm, leveraging deep learning techniques, efficiently and accurately identifies patients with significant CAD using ExECG data, aiding clinical screening in both symptomatic and asymptomatic patients. Nevertheless, the specificity remains moderate (0.60), suggesting a potential for false positives and highlighting the need for further investigation.<\/jats:p><\/jats:sec>","DOI":"10.3389\/frai.2025.1496109","type":"journal-article","created":{"date-parts":[[2025,3,17]],"date-time":"2025-03-17T07:23:31Z","timestamp":1742196211000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease"],"prefix":"10.3389","volume":"8","author":[{"given":"Hsin-Yueh","family":"Liang","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kai-Cheng","family":"Hsu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shang-Yu","family":"Chien","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Chen-Yu","family":"Yeh","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ting-Hsuan","family":"Sun","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Meng-Hsuan","family":"Liu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kee Koon","family":"Ng","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2025,3,17]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"e008437","DOI":"10.1161\/CIRCEP.120.008437","article-title":"Artificial intelligence-enabled ECG algorithm to identify patients with left ventricular systolic dysfunction presenting to the emergency department with dyspnea","volume":"13","author":"Adedinsewo","year":"2020","journal-title":"Circ. Arrhythm. Electrophysiol."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"346","DOI":"10.1016\/j.mayocp.2014.12.013","article-title":"Maximal exercise testing variables and 10-year survival: fitness risk score derivation from the FIT project","volume":"90","author":"Ahmed","year":"2015","journal-title":"Mayo Clin. Proc."},{"key":"ref3","doi-asserted-by":"publisher","first-page":"e12879","DOI":"10.1111\/exsy.12879","article-title":"Optimal multiple key-based homomorphic encryption with deep neural networks to secure medical data transmission and diagnosis","volume":"39","author":"Alzubi","year":"2021","journal-title":"Expert. Syst."},{"key":"ref4","doi-asserted-by":"publisher","first-page":"649","DOI":"10.1067\/mhj.2002.124869","article-title":"Impact of diabetes mellitus on epicardial and microvascular flow after fibrinolytic therapy","volume":"144","author":"Angeja","year":"2002","journal-title":"Am. Heart J."},{"key":"ref5","doi-asserted-by":"publisher","first-page":"191","DOI":"10.1016\/j.jcmg.2010.10.011","article-title":"Quantification of coronary arterial stenoses by multidetector CT angiography in comparison with conventional angiography methods, caveats, and implications","volume":"4","author":"Arbab-Zadeh","year":"2011","journal-title":"JACC Cardiovasc. Imaging"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"861","DOI":"10.1016\/S0140-6736(19)31721-0","article-title":"An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction","volume":"394","author":"Attia","year":"2019","journal-title":"Lancet"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"2562","DOI":"10.1016\/j.eswa.2007.11.013","article-title":"Assessment of exercise stress testing with artificial neural network in determining coronary artery disease and predicting lesion localization","volume":"36","author":"Babaoglu","year":"2009","journal-title":"Expert Syst. Appl."},{"key":"ref8","doi-asserted-by":"publisher","first-page":"2182","DOI":"10.1016\/j.eswa.2009.07.055","article-title":"Effects of principle component analysis on assessment of coronary artery diseases using support vector machine","volume":"37","author":"Babao\u011flu","year":"2010","journal-title":"Expert Syst. Appl."},{"key":"ref9","author":"Banerjee","year":"2020"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"e3001627","DOI":"10.1371\/journal.pbio.3001627","article-title":"Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging","volume":"20","author":"Benkarim","year":"2022","journal-title":"PLoS Biol."},{"key":"ref11","volume-title":"Cardiovascular Intervention: a companion to Braunwald\u2019s Heart Disease","author":"Bhatt","year":"2015"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"1001982","DOI":"10.3389\/fcvm.2022.1001982","article-title":"Artificial intelligence-assisted remote detection of ST-elevation myocardial infarction using a mini-12-lead electrocardiogram device in prehospital ambulance care","volume":"9","author":"Chen","year":"2022","journal-title":"Front. Cardiovasc. Med."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"365","DOI":"10.1186\/s12911-021-01736-y","article-title":"ECG signal classification based on deep CNN and BiLSTM","volume":"21","author":"Cheng","year":"2021","journal-title":"BMC Med. Inform. Decis. Mak."},{"key":"ref14","doi-asserted-by":"publisher","first-page":"1264","DOI":"10.1016\/j.jacc.2013.11.052","article-title":"Yield of downstream tests after exercise treadmill testing: a prospective cohort study","volume":"63","author":"Christman","year":"2014","journal-title":"J. Am. Coll. Cardiol."},{"key":"ref15","doi-asserted-by":"publisher","first-page":"173","DOI":"10.1016\/0033-0620(89)90025-x","article-title":"The diagnostic accuracy of the exercise electrocardiogram: a meta-analysis of 22 years of research","volume":"32","author":"Detrano","year":"1989","journal-title":"Prog. Cardiovasc. Dis."},{"key":"ref16","doi-asserted-by":"publisher","first-page":"873","DOI":"10.1161\/CIR.0b013e31829b5b44","article-title":"Exercise standards for testing and training: a scientific statement from the American Heart Association","volume":"128","author":"Fletcher","year":"2013","journal-title":"Circulation"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"281","DOI":"10.1016\/s0167-5273(99)00054-6","article-title":"Exercise electrocardiography test in patients with aortic stenosis. Differential features from that of coronary artery disease","volume":"69","author":"Gencbay","year":"1999","journal-title":"Int. J. Cardiol."},{"key":"ref18","doi-asserted-by":"publisher","first-page":"e12370","DOI":"10.1111\/anec.12370","article-title":"Predictive value of exercise stress test-induced ST-segment changes in leads V(1) and avR in determining angiographic coronary involvement","volume":"22","author":"Ghaffari","year":"2017","journal-title":"Ann. Noninvasive Electrocardiol."},{"key":"ref19","doi-asserted-by":"publisher","first-page":"1531","DOI":"10.1016\/s0735-1097(02)02164-2","article-title":"ACC\/AHA 2002 guideline update for exercise testing: summary article. A report of the American College of Cardiology\/American Heart Association task force on practice guidelines (committee to update the 1997 exercise testing guidelines)","volume":"40","author":"Gibbons","year":"2002","journal-title":"J. Am. Coll. Cardiol."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"PP","DOI":"10.1109\/TNNLS.2022.3231917","article-title":"Neural network with a preference sampling paradigm for imbalanced data classification","author":"Huang","year":"2023","journal-title":"IEEE Trans. Neural. Netw. Learn Syst."},{"key":"ref21","doi-asserted-by":"publisher","first-page":"394","DOI":"10.3390\/biomedicines10020394","article-title":"An artificial intelligence-enabled ECG algorithm for the prediction and localization of angiography-proven coronary artery disease","volume":"10","author":"Huang","year":"2022","journal-title":"Biomedicines"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"2608","DOI":"10.1109\/EMBC.2015.7318926","article-title":"Convolutional neural networks for patient-specific ECG classification","volume":"2015","author":"Kiranyaz","year":"2015","journal-title":"Annu. Int. Conf. IEEE Eng. Med. Biol. Soc."},{"key":"ref23","doi-asserted-by":"publisher","first-page":"407","DOI":"10.1093\/eurheartj\/ehz425","article-title":"2019 ESC guidelines for the diagnosis and management of chronic coronary syndromes","volume":"41","author":"Knuuti","year":"2020","journal-title":"Eur. Heart J."},{"key":"ref24","doi-asserted-by":"crossref","DOI":"10.1007\/978-981-15-6325-6","volume-title":"Deep learning for medical decision support systems","author":"Kose","year":"2021"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"23","DOI":"10.1016\/j.atherosclerosis.2021.11.028","article-title":"Machine learning of treadmill exercise test to improve selection for testing for coronary artery disease","volume":"340","author":"Lee","year":"2022","journal-title":"Atherosclerosis"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"198","DOI":"10.1016\/s0022-0736(99)90080-8","article-title":"ST\/HR hysteresis: exercise and recovery phase ST depression\/heart rate analysis of the exercise ECG","volume":"32","author":"Lehtinen","year":"1999","journal-title":"J. Electrocardiol."},{"key":"ref27","doi-asserted-by":"publisher","first-page":"16812","DOI":"10.1002\/jcp.28350","article-title":"A review on coronary artery disease, its risk factors, and therapeutics","volume":"234","author":"Malakar","year":"2019","journal-title":"J. Cell. Physiol."},{"key":"ref28","doi-asserted-by":"publisher","first-page":"1180","DOI":"10.1016\/0735-1097(96)81472-0","article-title":"Influence of left ventricular hypertrophy on detection of coronary artery disease using exercise echocardiography","volume":"26","author":"Marwick","year":"1995","journal-title":"J. Am. Coll. Cardiol."},{"key":"ref29","doi-asserted-by":"publisher","first-page":"294","DOI":"10.4037\/aacnacc2019182","article-title":"Chronotropic incompetence","volume":"30","author":"Marzlin","year":"2019","journal-title":"AACN Adv. Crit. Care"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"89","DOI":"10.1038\/s41586-019-1799-6","article-title":"International evaluation of an AI system for breast cancer screening","volume":"577","author":"McKinney","year":"2020","journal-title":"Nature"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"350","DOI":"10.1161\/CIR.0000000000000061","article-title":"Role of noninvasive testing in the clinical evaluation of women with suspected ischemic heart disease: a consensus statement from the American Heart Association","volume":"130","author":"Mieres","year":"2014","journal-title":"Circulation"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"6017","DOI":"10.1007\/s12652-020-02623-6","article-title":"Artificial neural networks training algorithm integrating invasive weed optimization with differential evolutionary model","volume":"14","author":"Movassagh","year":"2023","journal-title":"J. Ambient. Intell. Humaniz. Comput."},{"key":"ref33","doi-asserted-by":"publisher","first-page":"1793","DOI":"10.1161\/CIRCULATIONAHA.113.001952","article-title":"Comparison of clinical interpretation with visual assessment and quantitative coronary angiography in patients undergoing percutaneous coronary intervention in contemporary practice: the assessing angiography (A2) project","volume":"127","author":"Nallamothu","year":"2013","journal-title":"Circulation"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"434","DOI":"10.2337\/diabetes.53.2.434","article-title":"Electrocardiographic repolarization complexity and abnormality predict all-cause and cardiovascular mortality in diabetes: the strong heart study","volume":"53","author":"Okin","year":"2004","journal-title":"Diabetes"},{"key":"ref35","doi-asserted-by":"publisher","first-page":"313","DOI":"10.1007\/s40618-016-0564-z","article-title":"Predictors of electrocardiographic abnormalities in type 1 diabetes: the Wisconsin epidemiologic study of diabetic retinopathy","volume":"40","author":"O'Neal","year":"2017","journal-title":"J. Endocrinol. Investig."},{"key":"ref36","doi-asserted-by":"publisher","first-page":"23","DOI":"10.1007\/s11883-022-00980-5","article-title":"Sex differences in coronary atherosclerosis","volume":"24","author":"Sato","year":"2022","journal-title":"Curr. Atheroscler. Rep."},{"key":"ref37","doi-asserted-by":"publisher","first-page":"89","DOI":"10.1007\/s11906-017-0787-1","article-title":"Blood pressure response to exercise and cardiovascular disease","volume":"19","author":"Schultz","year":"2017","journal-title":"Curr. Hypertens. Rep."},{"key":"ref38","doi-asserted-by":"publisher","first-page":"1622","DOI":"10.1161\/01.CIR.98.16.1622","article-title":"Use of a prognostic treadmill score in identifying diagnostic coronary disease subgroups","volume":"98","author":"Shaw","year":"1998","journal-title":"Circulation"},{"key":"ref39","doi-asserted-by":"publisher","first-page":"e23044","DOI":"10.1371\/journal.pone.0023044","article-title":"The accuracy of the electrocardiogram during exercise stress test based on heart size","volume":"6","author":"Siegler","year":"2011","journal-title":"PLoS One"},{"key":"ref40","doi-asserted-by":"publisher","first-page":"641","DOI":"10.1016\/s0735-1097(97)00217-9","article-title":"Importance of estimated functional capacity as a predictor of all-cause mortality among patients referred for exercise thallium single-photon emission computed tomography: report of 3,400 patients from a single center","volume":"30","author":"Snader","year":"1997","journal-title":"J. Am. Coll. Cardiol."},{"key":"ref41","first-page":"2361","volume-title":"The global burden of cardiovascular diseases and risk: a compass for future health","author":"Vaduganathan","year":"2022"},{"key":"ref42","author":"Verma","year":"2018"},{"key":"ref43","doi-asserted-by":"publisher","first-page":"3415","DOI":"10.1093\/eurheartj\/ehae177","article-title":"2024 ESC guidelines for the management of chronic coronary syndromes","volume":"45","author":"Vrints","year":"2024","journal-title":"Eur. Heart J."},{"key":"ref44","doi-asserted-by":"publisher","first-page":"113","DOI":"10.1186\/s40537-019-0276-2","article-title":"Deep convolutional neural network based medical image classification for disease diagnosis","volume":"6","author":"Yadav","year":"2019","journal-title":"J. Big Data"},{"key":"ref45","author":"Yang","year":"2016"},{"key":"ref46","doi-asserted-by":"publisher","first-page":"411","DOI":"10.1016\/j.compbiomed.2018.09.009","article-title":"Arrhythmia detection using deep convolutional neural network with long duration ECG signals","volume":"102","author":"Yildirim","year":"2018","journal-title":"Comput. Biol. Med."},{"key":"ref47","doi-asserted-by":"publisher","first-page":"101482","DOI":"10.1016\/j.cpcardiol.2022.101482","article-title":"Machine learning approach on high risk treadmill exercise test to predict obstructive coronary artery disease by using P, QRS, and T waves' features","volume":"48","author":"Yilmaz","year":"2023","journal-title":"Curr. Probl. Cardiol."},{"key":"ref48","doi-asserted-by":"publisher","first-page":"101856","DOI":"10.1016\/j.artmed.2020.101856","article-title":"ECG-based multi-class arrhythmia detection using spatio-temporal attention-based convolutional recurrent neural network","volume":"106","author":"Zhang","year":"2020","journal-title":"Artif. Intell. Med."},{"key":"ref49","author":"Zihlmann","year":"2017"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1496109\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,17]],"date-time":"2025-03-17T07:23:34Z","timestamp":1742196214000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1496109\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,17]]},"references-count":49,"alternative-id":["10.3389\/frai.2025.1496109"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1496109","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,3,17]]},"article-number":"1496109"}}