基于人工智能的心电全景图：思想、方法与应用前瞻

詹泽汇; 陈涵; 陈晋泰

doi:null

您当前的位置：

首页 >

文章列表页 >

基于人工智能的心电全景图：思想、方法与应用前瞻

特约专稿 | 更新时间：2026-01-06

- 基于人工智能的心电全景图：思想、方法与应用前瞻
- Artificial-Intelligence-Generated Electrocardio Panorama: Concept, Methods and Prospects
- 临床心电学杂志 2025年34卷第6期页码：414-417
- 作者机构：
  
  1.511400 广东广州，香港科技大学（广州），信息枢纽，人工智能学域（詹泽汇，陈晋泰）
  2.510275 广东广州，中山大学生物医学工程学院（詹泽汇）
  3.310009 浙江杭州，浙江大学医学院附属第二医院心血管内科（陈涵）
- 作者简介：
  
  [ "陈晋泰，现为香港科技大学（广州）助理教授、博士生导师兼副研究员。博士毕业于浙江大学计算机学院，曾在美国伊利诺伊大学香槟分校（University of Illinois Urbana-Champaign，UIUC）从事博士后研究。陈博士长期从事医学人工智能领域研究，主要研究方向包括医疗多模态大模型、结构化数据基础模型、智能药物设计与开发以及心脏病人工智能等。在Nature Communications、ICML、ICLR、NeurIPS等国际顶级期刊与会议发表论文七十余篇。科研成果曾获ICME最佳论文候选奖（前0.4%），在机器学习顶级国际会议口头报告六次、聚光灯展示一次，并获得中国电子学会科技进步二等奖、中国产学研合作促进会产学研合作创新成果二等奖、2025世界人工智能大会“云帆奖·璀璨明星”提名奖等荣誉。曾担任NeurIPS2025领域主席并任Nature Partner Journal Cardiovascular Health编委。 " ]
- 基金信息：
- DOI：
  中图分类号： R541.7;R540.4
- 收稿：2025-11-01，
  
  录用：2025-11-21，
  
  纸质出版：2025-12-28
- 稿件说明：
移动端阅览
詹泽汇, 陈涵, 陈晋泰. 基于人工智能的心电全景图：思想、方法与应用前瞻[J]. 临床心电学杂志, 2025, 34(6): 414-417.

ZHAN Zehui,CHEN Han,CHEN Jintai.Artificial-Intelligence-Generated Electrocardio Panorama: Concept, Methods and Prospects[J].J Clin Electrocardiol,2025,34(06):414-417.
詹泽汇, 陈涵, 陈晋泰. 基于人工智能的心电全景图：思想、方法与应用前瞻[J]. 临床心电学杂志, 2025, 34(6): 414-417. DOI：

ZHAN Zehui,CHEN Han,CHEN Jintai.Artificial-Intelligence-Generated Electrocardio Panorama: Concept, Methods and Prospects[J].J Clin Electrocardiol,2025,34(06):414-417. DOI：

摘要

传统12导联心电图在临床诊断中发挥着不可替代的作用，但其固定的导联配置限制了对心脏电活动进行多角度、全方位审视的能力。本文旨在系统阐述一种新兴技术——“心电全景图”的基本概念、技术原理及其临床转化潜力。心电全景图的核心思想是：利用人工智能模型，从有限物理导联信号中重建心脏电活动向量场，进而合成任意虚拟视角下的高保真心电图。本文以左、右束支传导阻滞为例，初步验证了该技术的有效性。结果显示，基于Ⅰ、Ⅱ、V

导联合成的V

、V

导联信号，在关键诊断特征（如LBBB的宽大QS波、RBBB的“兔耳征”）上与真实信号高度一致；在进一步生成的多组非标准虚拟视角心电图中，疾病特异性波形仍能稳定呈现。这一结果表明，该技术突破了传统导联的物理限制，实现了对心脏电活动的全景化、动态化观测。

Abstract

The conventional 12-lead electrocardiogram plays an essential role in clinical diagnosis

yet its fixed lead configuration restricts the ability to observe cardiac electrical activity from diverse and comprehensive viewpoints. This article presents a systematic overview of an emerging technology known as

Electrocardio Panorama

outlining its conceptual foundation

technical framework and potential for clinical translation. The central idea is to use artificial intelligence generative models to reconstruct the vector field of cardiac electrical activity from a limited set of physical leads and to synthesize high-fidelity electrocardiograms from any virtual viewpoint. Using left and right bundle branch block as examples

we performed a preliminary evaluation of this approach. Signals synthesize

d for leads V

and V

from leads Ⅰ

Ⅱ and V

reproduced key diagnostic characteristics with high fidelity

such as the broad QS complex typical of left bundle branch block and the “rabbit-ear” pattern associated with right bundle branch block. In the additional non-standard virtual viewpoints generated

disease-specific waveforms remained consistently identifiable. These findings demonstrate that this technique transcends the physical constraints of traditional lead systems and enables panoramic

dynamic visualization of cardiac electrical activity.

关键词

Keywords

references

HOLTER N J . New method for heart studies [J ] . Noninvasive Electrocardiol , 1998 , 3 ( 4 ): 381 - 387 .

VAN GORSELEN E O F , VERHEUGT F W A , MEURSING B T J , et al . Posterior myocardial infarction: the dark side of the moon [J ] . Neth Heart J , 2007 , 15 ( 1 ): 16 - 21 .

BERNE P , BRUGADA J . Brugada syndrome 2012 [J ] . Circ J , 2012 , 76 ( 7 ): 1563 - 1571 .

DREW B J , CALIFF R M , FUNK M , et al . Practice standards for electrocardiographic monitoring in hospital settings: an American Heart Association scientific statement from the Councils on Cardiovascular Nursing,Clinical Cardiology,and Cardiovascular Disease in the Young: endorsed by the International Society of Computerized Electrocardiology and the American Association of Critical-Care Nurses [J ] . Circulation , 2004 , 110 ( 17 ): 2721 - 2746 .

CHEN J , ZHENG X , YU H , et al . Electrocardio Panorama: Synthesizing New ECG Views with Self-supervision [J ] . In Proceedings of the International Joint Conference on Artificial Intelligence 2021 Aug.arXiv preprint arXiv: 2105. 06293

Mildenhall B , Srinivasan PP , Tancik M , Barron JT , Ramamoorthi R , Ng R . Nerf: Representing scenes as neural radiance fields for view synthesis [J ] . Communications of the ACM , 2021 , 65 ( 1 ): 99 - 106 .

GRANT R P . Spatial vector electrocardiography;a method for calculating the spatial electrical vectors of the heart from conventional leads [J ] . Circulation , 1950 , 2 ( 5 ): 676 - 695 ..

ZEHUI ZHAN , YAOJUN HU , JIAJING ZHAN , et al . NEF-NET+: Adapting Electrocardio panorama in the wild [J ] . arXiv preprint arXiv: 2511. 02880 . 2025 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

人工智能在ECG解读与临床应用