心电特征引导下的自监督房颤异常检测方法

doi:10.3778/j.issn.1002-8331.2309-0166

摘要/Abstract

摘要： 心电异常检测旨在发现心电数据中异常的模式，例如房颤特征或无效干扰信号特征。针对心房颤动心电异常，提出了一种简单有效的自监督房颤异常检测方法，称为心电特征引导下的房颤异常检测器（electrocardiogram feature induced atrial fibrillation detector，EFAFD），通过引入P波掩码和心率变异性指标预测多任务学习策略，指导模型学习房颤心电P波消失和RR间期绝对不齐等医学特征，提高模型对房颤异常心电模式的判别能力。具体地，将P波掩码心电数据通过自编码器重构原始的心电数据，学习房颤心电P波易消失的特征。同时，将心率变异性指标的预测任务整合到自编码器框架中，学习房颤心电RR间期绝对不齐的节律特征。通过度量心电的重构误差，实现房颤心电的检测。在真实的动态心电数据集上评估了所提出的方法，包括CPSC2021数据集和Icentia11k数据集。EFAFD模型的AUC分别达到了81.85%和92.46%。实验结果表明，所提出的方法在房颤异常检测方面优于现有的方法。

关键词: 心电图, 时间序列, 异常检测, 自监督学习, 房颤检测, 自编码器

Abstract: Electrocardiogram (ECG) anomaly detection aims to find abnormal patterns in ECG data, such as atrial fibrillation features or invalid interference signal features. A simple and effective self-supervised anomaly detection method has been proposed for atrial fibrillation ECG disease, called electrocardiogram feature induced atrial fibrillation detector (EFAFD). In this method, the P-wave mask and heart rate variability (HRV) index are introduced through a multi-task learning strategy to guide the model in learning the medical characteristics such as P-wave disappearance and irregularity of RR intervals, so as to improve the model’s ability to distinguish abnormal ECG patterns of atrial fibrillation. Specifically, the ECG data with P-wave masking are reconstructed using an autoencoder to capture the characteristic of P-wave disappearance in atrial fibrillation. Simultaneously, the prediction task of HRV indices is integrated into the autoencoder framework to learn the rhythmic features of atrial fibrillation electrocardiogram with absolute irregularity of RR intervals. Finally, atrial fibrillation ECG is detected by measuring the reconstruction error of the ECG. The proposed method is evaluated on real dynamic ECG datasets, including the CPSC2021 dataset and the Icentia11k dataset. The AUC of EFAFD reaches 81.85% on the CPSC2021 dataset and 92.46% on the Icentia11k dataset, and experimental results demonstrate its superiority over existing methods in atrial fibrillation anomaly detection.

Key words: electrocardiogram, time series, anomaly detection, self-supervised learning, atrial fibrillation detection, autoencoder

陈鹏, 邓淼磊, 樊好义, 张德贤, 韩涵. 心电特征引导下的自监督房颤异常检测方法[J]. 计算机工程与应用, 2025, 61(2): 208-218.

CHEN Peng, DENG Miaolei, FAN Haoyi, ZHANG Dexian, HAN Han. Self-Supervised Atrial Fibrillation Anomaly Detection Method Guided by Electrocardiogram Features[J]. Computer Engineering and Applications, 2025, 61(2): 208-218.

参考文献

[1] DEWI W N, SAFRI S, ROSMA I H. Modified precordial lead ECG SafOne on electrocardiography recordings[J]. Scientific Reports, 2022, 12(1): 7934.
[2] ZONI-BERISSO M, LERCARI F, CARAZZA T, et al. Epidemiology of atrial fibrillation: European perspective[J]. Clinical Epidemiology, 2014(6): 213-220.
[3] KORNEJ J, BORSCHEL C S, BENJAMIN E J, et al. Epidemiology of atrial fibrillation in the 21st century: novel methods and new insights[J]. Circulation Research, 2020, 127(1): 4-20.
[4] ZHANG R, ISOLA P, EFROS A A. Split-brain autoencoders: unsupervised learning by cross-channel prediction[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 1058-1067.
[5] KIYASSEH D, ZHU T, CLIFTON D A. Clocs: contrastive learning of cardiac signals across space, time, and patients[C]//Proceedings of the International Conference on Machine Learning, 2021: 5606-5615.
[6] OH J, CHUNG H, KWON J, et al. Lead-agnostic self-supervised learning for local and global representations of electrocardiogram[C]//Proceedings of the Conference on Health, Inference, and Learning, 2022: 338-353.
[7] MEHARI T, STRODTHOFF N. Self-supervised representation learning from 12-lead ECG data[J]. Computers in Biology and Medicine, 2022, 141: 105114.
[8] 陈健, 刘明, 熊鹏, 等. 基于卷积自编码神经网络的心电信号降噪[J]. 计算机工程与应用, 2020, 56(16): 148-155.
CHEN J, LIU M, XIONG P, et al. ECG signal denoising based on convolutional auto-encoder neural network[J]. Computer Engineering and Applications, 2020, 56(16): 148-155.
[9] ZHOU C, PAFFENROTH R C. Anomaly detection with robust deep autoencoders[C]//Proceedings of the 23rd International Conference on Knowledge Discovery and Data Mining, 2017: 665-674.
[10] RUFF L, VANDERMEULEN R, GOERNITZ N, et al. Deep one-class classification[C]//Proceedings of the International Conference on Machine Learning, 2018: 4393-4402.
[11] LI Z, ZHAO Y, HU X, et al. ECOD: unsupervised outlier detection using empirical cumulative distribution functions[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(12): 12181-12193.
[12] YUE H G, JUN H, XU L, et al. Effectiveness of single-lead ECG devices for detecting atrial fibrillation: an overview of systematic reviews[J]. Worldviews on Evidence-Based Nursing, 2024, 21(1): 79-86.
[13] SERMANET P, LYNCH C, HSU J, et al. Time-contrastive networks: self-supervised learning from multi-view observation[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2017: 486-487.
[14] FRANCESCHI J Y, DIEULEVEUT A, JAGGI M. Unsupervised scalable representation learning for multivariate time series[C]//Proceedings of the 33rd International Conference on Neural Information Processing Systems, 2019: 4650-4661.
[15] SCHNEIDER S, BAEVSKI A, COLLOBERT R, et al. wav2vec: unsupervised pre-training for speech recognition[J]. arXiv:1904.05862, 2019.
[16] YèCHE H, DRESDNER G, LOCATELLO F, et al. Neighborhood contrastive learning applied to online patient monitoring[C]//Proceedings of the International Conference on Machine Learning, 2021: 11964-11974.
[17] PASCUAL S, RAVANELLI M, SERRA J, et al. Learning problem-agnostic speech representations from multiple self-supervised tasks[J]. arXiv:1904.03416, 2019.
[18] RAVANELLI M, ZHONG J, PASCUAL S, et al. Multi-task self-supervised learning for robust speech recognition[C]//Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, 2020: 6989-6993.
[19] SAEED A, OZCELEBI T, LUKKIEN J. Multi-task self-supervised learning for human activity detection[J]. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 2019, 3(2): 1-30.
[20] SARKAR P, ETEMAD A. Self-supervised learning for ECG-based emotion recognition[C]//Proceedings of the 2020 IEEE International Conference on Acoustics, Speech and Signal Processing, 2020: 3217-3221.
[21] MIAO Q, XU C, ZHAN J, et al. An unsupervised short-and long-term mask representation for multivariate time series anomaly detection[C]//Proceedings of the International Conference on Neural Information Processing, 2022: 504-516.
[22] ZONG B, SONG Q, MIN M R, et al. Deep autoencoding Gaussian mixture model for unsupervised anomaly detection[C]//Proceedings of the International Conference on Learning Representations, 2018.
[23] ZHANG C, SONG D, CHEN Y, et al. A deep neural network for unsupervised anomaly detection and diagnosis in multivariate time series data[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2019: 1409-1416.
[24] PARK D, HOSHI Y, KEMP C C. A multimodal anomaly detector for robot-assisted feeding using an LSTM-based variational autoencoder[J]. IEEE Robotics and Automation Letters, 2018, 3(3): 1544-1551.
[25] AUDIBERT J, MICHIARDI P, GUYARD F, et al. USAD: unsupervised anomaly detection on multivariate time series[C]//Proceedings of the 26th International Conference on Knowledge Discovery & Data Mining, 2020: 3395-3404.
[26] 王昊天, 郑栋毅, 刘芳, 等. 面向多元时序数据的个性化联邦异常检测方法[J]. 计算机工程与应用, 2022, 58(11): 60-65.
WANG H T, ZHENG D Y, LIU F, et al. Personalized federated anomaly detection method for multivariate time series data[J]. Computer Engineering and Applications, 2022, 58(11): 60-65.
[27] 席亮, 刘涵, 樊好义, 等. 基于深度对抗学习潜在表示分布的异常检测模型[J]. 电子学报, 2021, 49(7): 1257-1265.
Ⅺ L, LIU H, FAN H Y, et al. Deep adversarial learning latent representation distribution model for anomaly detection[J]. Acta Electronica Sinica, 2021, 49(7): 1257-1265.
[28] 席亮, 王瑞东, 樊好义, 等. 基于样本关联感知的无监督深度异常检测模型[J]. 计算机学报, 2021, 44(11): 2317-2331.
Ⅺ L, WANG R D, FAN H Y. Sample-correlation-aware unsupervised deep anomaly detection model[J]. Chinese Journal of Computers, 2021, 44(11): 2317-2331.
[29] 黄训华, 张凤斌, 樊好义, 等. 基于多模态对抗学习的无监督时间序列异常检测[J]. 计算机研究与发展, 2021, 58(8): 1655-1667.
HUANG X H, ZHANG F B, FAN H Y, et al. Multimodal adversarial learning based unsupervised time series anomaly detection[J]. Journal of Computer Research and Development, 2021, 58(8): 1655-1667.
[30] SU Y, ZHAO Y, NIU C, et al. Robust anomaly detection for multivariate time series through stochastic recurrent neural network[C]//Proceedings of the 25th International Conference on Knowledge Discovery & Data Mining, 2019: 2828-2837.
[31] LI Z, ZHAO Y, HAN J, et al. Multivariate time series anomaly detection and interpretation using hierarchical inter-metric and temporal embedding[C]//Proceedings of the 27th Conference on Knowledge Discovery & Data Mining, 2021: 3220-3230.
[32] SILIPO R, MARCHESI C. Artificial neural networks for automatic ECG analysis[J]. IEEE Transactions on Signal Processing, 1998, 46(5): 1417-1425.
[33] LAGERHOLM M, PETERSON C, BRACCINI G, et al. Clustering ECG complexes using Hermite functions and self-organizing maps[J]. IEEE Transactions on Biomedical Engineering, 2000, 47(7): 838-848.
[34] BENALI R, REGUIG B F, SLIMANEHZ. Automatic classification of heartbeats using wavelet neural network[J]. Journal of Medical Systems, 2012, 36: 883-892.
[35] ACIR N. Classification of ECG beats by using a fast least square support vector machines with a dynamic programming feature selection algorithm[J]. Neural Computing & Applications, 2005, 14: 299-309.
[36] KAO W C, YU C K, SHEN C P, et al. Electrocardiogram analysis with adaptive feature selection and support vector machines[C]//Proceedings of the 2006 IEEE Asia Pacific Conference on Circuits and Systems, 2006: 1783-1786.
[37] SCHLEGL T, SEEB?CK P, WALDSTEIN S M, et al. Unsupervised anomaly detection with generative adversarial networks to guide marker discovery[C]//Proceedings of the International Conference on Information Processing in Medical Imaging, 2017: 146-157.
[38] AKCAY S, ATAPOUR-ABARGHOUEI A, BRECKON T P. GANomaly: semi-supervised anomaly detection via adversarial training[C]//Proceedings of the 14th Asian Conference on Computer Vision, 2019: 622-637.
[39] 庄跃生, 林珊玲, 林志贤, 等 . 生成对抗网络在数据异常检测中的研究[J]. 计算机工程与应用, 2022, 58(4): 143-149.
ZHUANG Y S, LIN S L, LIN Z X, et al. Study on generative adversarial network for data anomaly detection[J]. Computer Engineering and Applications, 2022, 58(4): 143-149.
[40] ZHOU B, LIU S, HOOI B, et al. BeatGAN: anomalous rhythm detection using adversarially generated time series[C]//Proceedings of the International Joint Conference on Artificial Intelligence, 2019: 4433-4439.
[41] 王润极. 应用心率变异性指标监控运动训练的研究进展[J]. 福建体育科技, 2023, 42(4): 94-98.
WANG R J. Advances in the application of heart rate variability indices to monitor exercise training[J]. Fujian Sports Science and Technology, 2023, 42(4): 94-98.
[42] XU H, WANG Y, JIAN S, et al. Calibrated one-class classification for unsupervised time series anomaly detection[J]. arXiv:2207.12201, 2022.
[43] WANG R, LIU C, MOU X, et al. Deep contrastive one-class time series anomaly detection[C]//Proceedings of the International Conference on Data Mining, 2023: 694-702.