Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease

BackgroundThe 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).MethodsWe retrospectively collected 818 p...

Full description

Saved in:
Bibliographic Details
Main Authors: Hsin-Yueh Liang, Kai-Cheng Hsu, Shang-Yu Chien, Chen-Yu Yeh, Ting-Hsuan Sun, Meng-Hsuan Liu, Kee Koon Ng
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-03-01
Series:Frontiers in Artificial Intelligence
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/frai.2025.1496109/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850040838214647808
author Hsin-Yueh Liang
Hsin-Yueh Liang
Kai-Cheng Hsu
Kai-Cheng Hsu
Kai-Cheng Hsu
Shang-Yu Chien
Chen-Yu Yeh
Ting-Hsuan Sun
Meng-Hsuan Liu
Kee Koon Ng
author_facet Hsin-Yueh Liang
Hsin-Yueh Liang
Kai-Cheng Hsu
Kai-Cheng Hsu
Kai-Cheng Hsu
Shang-Yu Chien
Chen-Yu Yeh
Ting-Hsuan Sun
Meng-Hsuan Liu
Kee Koon Ng
author_sort Hsin-Yueh Liang
collection DOAJ
description BackgroundThe 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).MethodsWe retrospectively collected 818 patients who underwent both ExECG and coronary angiography (CAG) within 6 months. The mean age was 57.0 ± 10.1 years, 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.ResultsUsing 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.ConclusionThe 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.
format Article
id doaj-art-d2f9461609ea4c49bfca238541f804e1
institution DOAJ
issn 2624-8212
language English
publishDate 2025-03-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Artificial Intelligence
spelling doaj-art-d2f9461609ea4c49bfca238541f804e12025-08-20T02:55:57ZengFrontiers Media S.A.Frontiers in Artificial Intelligence2624-82122025-03-01810.3389/frai.2025.14961091496109Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery diseaseHsin-Yueh Liang0Hsin-Yueh Liang1Kai-Cheng Hsu2Kai-Cheng Hsu3Kai-Cheng Hsu4Shang-Yu Chien5Chen-Yu Yeh6Ting-Hsuan Sun7Meng-Hsuan Liu8Kee Koon Ng9Division of Cardiology, Department of Medicine, China Medical University Hospital, Taichung, TaiwanDepartment of Biomedical Imaging and Radiological Science, China Medical University, Taichung, TaiwanArtificial Intelligence Center, China Medical University Hospital, Taichung, TaiwanSchool of Medicine, China Medical University, Taichung, TaiwanDepartment of Neurology, China Medical University Hospital, Taichung, TaiwanArtificial Intelligence Center, China Medical University Hospital, Taichung, TaiwanArtificial Intelligence Center, China Medical University Hospital, Taichung, TaiwanArtificial Intelligence Center, China Medical University Hospital, Taichung, TaiwanArtificial Intelligence Center, China Medical University Hospital, Taichung, TaiwanDivision of Cardiology, Department of Medicine, China Medical University Hospital, Taichung, TaiwanBackgroundThe 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).MethodsWe retrospectively collected 818 patients who underwent both ExECG and coronary angiography (CAG) within 6 months. The mean age was 57.0 ± 10.1 years, 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.ResultsUsing 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.ConclusionThe 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.https://www.frontiersin.org/articles/10.3389/frai.2025.1496109/fullexercise stress electrocardiographycoronary artery diseasedeep learningmultimodal approachfeature variableartificial intelligence
spellingShingle Hsin-Yueh Liang
Hsin-Yueh Liang
Kai-Cheng Hsu
Kai-Cheng Hsu
Kai-Cheng Hsu
Shang-Yu Chien
Chen-Yu Yeh
Ting-Hsuan Sun
Meng-Hsuan Liu
Kee Koon Ng
Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease
Frontiers in Artificial Intelligence
exercise stress electrocardiography
coronary artery disease
deep learning
multimodal approach
feature variable
artificial intelligence
title Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease
title_full Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease
title_fullStr Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease
title_full_unstemmed Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease
title_short Deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease
title_sort deep learning analysis of exercise stress electrocardiography for identification of significant coronary artery disease
topic exercise stress electrocardiography
coronary artery disease
deep learning
multimodal approach
feature variable
artificial intelligence
url https://www.frontiersin.org/articles/10.3389/frai.2025.1496109/full
work_keys_str_mv AT hsinyuehliang deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT hsinyuehliang deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT kaichenghsu deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT kaichenghsu deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT kaichenghsu deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT shangyuchien deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT chenyuyeh deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT tinghsuansun deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT menghsuanliu deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease
AT keekoonng deeplearninganalysisofexercisestresselectrocardiographyforidentificationofsignificantcoronaryarterydisease