Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantation

BackgroundEffective management of dual antiplatelet therapy (DAPT) following drug-eluting stent (DES) implantation is crucial for preventing adverse events. Traditional prognostic tools, such as rule-based methods or Cox regression, despite their widespread use and ease, tend to yield moderate predi...

Full description

Saved in:
Bibliographic Details
Main Authors: Fang Li, Zenan Sun, Ahmed abdelhameed, Tiehang Duan, Laila Rasmy, Xinyue Hu, Jianping He, Yifang Dang, Jingna Feng, Jianfu Li, Yichen Wang, Tianchen Lyu, Naomi Braun, Si Pham, Michael Gharacholou, DeLisa Fairweather, Degui Zhi, Jiang Bian, Cui Tao
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-01-01
Series:Frontiers in Cardiovascular Medicine
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fcvm.2024.1460354/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850098465605943296
author Fang Li
Zenan Sun
Ahmed abdelhameed
Tiehang Duan
Laila Rasmy
Xinyue Hu
Jianping He
Yifang Dang
Jingna Feng
Jianfu Li
Yichen Wang
Tianchen Lyu
Naomi Braun
Si Pham
Michael Gharacholou
DeLisa Fairweather
Degui Zhi
Jiang Bian
Cui Tao
author_facet Fang Li
Zenan Sun
Ahmed abdelhameed
Tiehang Duan
Laila Rasmy
Xinyue Hu
Jianping He
Yifang Dang
Jingna Feng
Jianfu Li
Yichen Wang
Tianchen Lyu
Naomi Braun
Si Pham
Michael Gharacholou
DeLisa Fairweather
Degui Zhi
Jiang Bian
Cui Tao
author_sort Fang Li
collection DOAJ
description BackgroundEffective management of dual antiplatelet therapy (DAPT) following drug-eluting stent (DES) implantation is crucial for preventing adverse events. Traditional prognostic tools, such as rule-based methods or Cox regression, despite their widespread use and ease, tend to yield moderate predictive accuracy within predetermined timeframes. This study introduces a new contrastive learning-based approach to enhance prediction efficacy over multiple time intervals.MethodsWe utilized retrospective, real-world data from the OneFlorida + Clinical Research Consortium. Our study focused on two primary endpoints: ischemic and bleeding events, with prediction windows of 1, 2, 3, 6, and 12 months post-DES implantation. Our approach first utilized an auto-encoder to compress patient features into a more manageable, condensed representation. Following this, we integrated a Transformer architecture with multi-head attention mechanisms to focus on and amplify the most salient features, optimizing the representation for better predictive accuracy. Then, we applied contrastive learning to enable the model to further refine its predictive capabilities by maximizing intra-class similarities and distinguishing inter-class differences. Meanwhile, the model was holistically optimized using multiple loss functions, to ensure the predicted results closely align with the ground-truth values from various perspectives. We benchmarked model performance against three cutting-edge deep learning-based survival models, i.e., DeepSurv, DeepHit, and SurvTrace.ResultsThe final cohort comprised 19,713 adult patients who underwent DES implantation with more than 1 month of records after coronary stenting. Our approach demonstrated superior predictive performance for both ischemic and bleeding events across prediction windows of 1, 2, 3, 6, and 12 months, with time-dependent concordance (Ctd) index values ranging from 0.88 to 0.80 and 0.82 to 0.77, respectively. It consistently outperformed the baseline models, including DeepSurv, DeepHit, and SurvTrace, with statistically significant improvement in the Ctd-index values for most evaluated scenarios.ConclusionThe robust performance of our contrastive learning-based model underscores its potential to enhance DAPT management significantly. By delivering precise predictive insights at multiple time points, our method meets the current need for adaptive, personalized therapeutic strategies in cardiology, thereby offering substantial value in improving patient outcomes.
format Article
id doaj-art-2c5fb60840c041c0b136994ba976ef9e
institution DOAJ
issn 2297-055X
language English
publishDate 2025-01-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Cardiovascular Medicine
spelling doaj-art-2c5fb60840c041c0b136994ba976ef9e2025-08-20T02:40:43ZengFrontiers Media S.A.Frontiers in Cardiovascular Medicine2297-055X2025-01-011110.3389/fcvm.2024.14603541460354Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantationFang Li0Zenan Sun1Ahmed abdelhameed2Tiehang Duan3Laila Rasmy4Xinyue Hu5Jianping He6Yifang Dang7Jingna Feng8Jianfu Li9Yichen Wang10Tianchen Lyu11Naomi Braun12Si Pham13Michael Gharacholou14DeLisa Fairweather15Degui Zhi16Jiang Bian17Cui Tao18Department of Artificial Intelligence and Informatics, Mayo Clinic, Jacksonville, FL, United StatesMcWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, TX, United StatesDepartment of Artificial Intelligence and Informatics, Mayo Clinic, Jacksonville, FL, United StatesDepartment of Artificial Intelligence and Informatics, Mayo Clinic, Jacksonville, FL, United StatesMcWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, TX, United StatesDepartment of Artificial Intelligence and Informatics, Mayo Clinic, Jacksonville, FL, United StatesMcWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, TX, United StatesMcWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, TX, United StatesDepartment of Artificial Intelligence and Informatics, Mayo Clinic, Jacksonville, FL, United StatesDepartment of Artificial Intelligence and Informatics, Mayo Clinic, Jacksonville, FL, United StatesDivision of Hospital Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United StatesHealth Outcomes and Biomedical Informatics, University of Florida Health, Gainesville, FL, United StatesHealth Outcomes and Biomedical Informatics, University of Florida Health, Gainesville, FL, United StatesDepartment of Cardiothoracic Surgery, Mayo Clinic, Jacksonville, FL, United StatesDepartment of Cardiovascular Medicine, Mayo Clinic, Jacksonville, FL, United StatesDepartment of Cardiovascular Medicine, Mayo Clinic, Jacksonville, FL, United StatesMcWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, TX, United StatesHealth Outcomes and Biomedical Informatics, University of Florida Health, Gainesville, FL, United StatesDepartment of Artificial Intelligence and Informatics, Mayo Clinic, Jacksonville, FL, United StatesBackgroundEffective management of dual antiplatelet therapy (DAPT) following drug-eluting stent (DES) implantation is crucial for preventing adverse events. Traditional prognostic tools, such as rule-based methods or Cox regression, despite their widespread use and ease, tend to yield moderate predictive accuracy within predetermined timeframes. This study introduces a new contrastive learning-based approach to enhance prediction efficacy over multiple time intervals.MethodsWe utilized retrospective, real-world data from the OneFlorida + Clinical Research Consortium. Our study focused on two primary endpoints: ischemic and bleeding events, with prediction windows of 1, 2, 3, 6, and 12 months post-DES implantation. Our approach first utilized an auto-encoder to compress patient features into a more manageable, condensed representation. Following this, we integrated a Transformer architecture with multi-head attention mechanisms to focus on and amplify the most salient features, optimizing the representation for better predictive accuracy. Then, we applied contrastive learning to enable the model to further refine its predictive capabilities by maximizing intra-class similarities and distinguishing inter-class differences. Meanwhile, the model was holistically optimized using multiple loss functions, to ensure the predicted results closely align with the ground-truth values from various perspectives. We benchmarked model performance against three cutting-edge deep learning-based survival models, i.e., DeepSurv, DeepHit, and SurvTrace.ResultsThe final cohort comprised 19,713 adult patients who underwent DES implantation with more than 1 month of records after coronary stenting. Our approach demonstrated superior predictive performance for both ischemic and bleeding events across prediction windows of 1, 2, 3, 6, and 12 months, with time-dependent concordance (Ctd) index values ranging from 0.88 to 0.80 and 0.82 to 0.77, respectively. It consistently outperformed the baseline models, including DeepSurv, DeepHit, and SurvTrace, with statistically significant improvement in the Ctd-index values for most evaluated scenarios.ConclusionThe robust performance of our contrastive learning-based model underscores its potential to enhance DAPT management significantly. By delivering precise predictive insights at multiple time points, our method meets the current need for adaptive, personalized therapeutic strategies in cardiology, thereby offering substantial value in improving patient outcomes.https://www.frontiersin.org/articles/10.3389/fcvm.2024.1460354/fulldual antiplatelet therapycontrastive learningtransformerpredictive modelingadverse endpointdrug-eluting coronary artery stent implantation
spellingShingle Fang Li
Zenan Sun
Ahmed abdelhameed
Tiehang Duan
Laila Rasmy
Xinyue Hu
Jianping He
Yifang Dang
Jingna Feng
Jianfu Li
Yichen Wang
Tianchen Lyu
Naomi Braun
Si Pham
Michael Gharacholou
DeLisa Fairweather
Degui Zhi
Jiang Bian
Cui Tao
Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantation
Frontiers in Cardiovascular Medicine
dual antiplatelet therapy
contrastive learning
transformer
predictive modeling
adverse endpoint
drug-eluting coronary artery stent implantation
title Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantation
title_full Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantation
title_fullStr Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantation
title_full_unstemmed Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantation
title_short Contrastive learning with transformer for adverse endpoint prediction in patients on DAPT post-coronary stent implantation
title_sort contrastive learning with transformer for adverse endpoint prediction in patients on dapt post coronary stent implantation
topic dual antiplatelet therapy
contrastive learning
transformer
predictive modeling
adverse endpoint
drug-eluting coronary artery stent implantation
url https://www.frontiersin.org/articles/10.3389/fcvm.2024.1460354/full
work_keys_str_mv AT fangli contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT zenansun contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT ahmedabdelhameed contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT tiehangduan contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT lailarasmy contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT xinyuehu contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT jianpinghe contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT yifangdang contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT jingnafeng contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT jianfuli contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT yichenwang contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT tianchenlyu contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT naomibraun contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT sipham contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT michaelgharacholou contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT delisafairweather contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT deguizhi contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT jiangbian contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation
AT cuitao contrastivelearningwithtransformerforadverseendpointpredictioninpatientsondaptpostcoronarystentimplantation