Abstract
Anomaly detection in electrocardiogram (ECG) signals is a crucial task in the medical domain. But determining a threshold for detecting the abnormal signal is challenging, and it significantly impacts the performance of a Deep Learning (DL) model. In this paper, we propose a methodology to assess the impact of variable thresholds on anomaly detection in ECG signals generated using the reconstruction error of an autoencoder model. First, we consider a set of random thresholds and generate the associated confusion matrix along with a ROC curve to analyze the varying performance. We then propose a method to systematically generate a set of thresholds and find the optimal threshold using the associated ROC curve. Our experiment suggests that the proposed method of using a ROC curve for a set of systematically generated thresholds to find the (near-) optimal threshold can potentially improve the detection of abnormalities in the ECG signal.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Weszka JS, Rosenfeld A (1978) Threshold evaluation techniques. IEEE Trans Syst Man Cybern 8(8):622–629
Fawcett T (2006) An introduction to roc analysis. Pattern Recogn Lett 27(8):861–874
Flach PA (2016) Roc analysis. In: Encyclopedia of machine learning and data mining. Springer, pp 1–8
Malhotra P, Ramakrishnan A, Anand G, Agarwal P, Shroff G (2016) LSTM-based encoder-decoder for multi-sensor anomaly detection, Lovekesh Vig
Chauhan S, Vig L (2015) Anomaly detection in ECG time signals via deep LSTM networks. In: Proceedings IEEE International conference on data science and advanced analytics (DSAA), pp 1–7
Cui KX, Xia XJ (2022) ECG signal anomaly detection algorithm based on CNN-BiLSTM. In: Proceedings 11th international conference information and communication technology (ICTech), pp 193–197
Savalia S, Emamian V (2018) Cardiac arrhythmia classification by multi-layer perceptron and convolution neural networks. Bioengineering 5(2)
Kieu T, Yang B, Guo C, Jensen CS (2019) Outlier detection for time series with recurrent autoencoder ensembles. In: Proceedings of 28th international joint conference artificial intelligence (IJCAI), pp 2725–2732
Wu X, Lu X, Leung H (2017) An adaptive threshold deep learning method for fire and smoke detection. In: Proceedings IEEE international conference systems, man, & cybernetics (SMC), pp 1954–1959
Hong CS (2009) Optimal threshold from roc and cap curves. Commun Stat-Simul Comput 38(10):2060–2072
Du S, Li T, Horng SJ (2018) Time series forecasting using sequence-to-sequence deep learning framework. In: Proceedings 9th International symposium parallel architectures, algorithms, and programming (PAAP), pp 171–176
Lai KH, Zha D, Xu J, Zhao Y, Wang G, Hu X (2021) Revisiting time series outlier detection: Definitions and benchmarks. In: Proceedings of 35th conference neural information processing systems datasets and benchmarks track (Round 1)
Blázquez-García A, Conde A, Mori U, Lozano JA (2021) A review on outlier/anomaly detection in time series data. ACM Comput Surv 54(3):1–33
Geiger A, Liu D, Alnegheimish S, Cuesta-Infante A, Veeramachaneni K (2020) TadGAN: time series anomaly detection using generative adversarial networks. In: Proceedings IEEE international conference big data, pp 33–43
Kao JB, Jiang JR (2019) Anomaly detection for univariate time series with statistics and deep learning. In: Proceedings IEEE Eurasia conference IOT, communication & engineering (ECICE), pp 404–407
Metz Charles E (1978) Basic principles of roc analysis. Semin Nucl Med 8(4):283–298
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Mishra, B., Kumar, R. (2023). Assessment of Variable Threshold for Anomaly Detection in ECG Time Signals with Deep Learning. In: Yadav, A., Nanda, S.J., Lim, MH. (eds) Proceedings of International Conference on Paradigms of Communication, Computing and Data Analytics. PCCDA 2023. Algorithms for Intelligent Systems. Springer, Singapore. https://doi.org/10.1007/978-981-99-4626-6_7
Download citation
DOI: https://doi.org/10.1007/978-981-99-4626-6_7
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-4625-9
Online ISBN: 978-981-99-4626-6
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)