Deep Learning Modeling of Cardiac Arrhythmia Classification on Information Feature Fusion Image with Attention Mechanism
Abstract
:1. Introduction
2. Data Processing Method
2.1. Introduction of the Dataset
2.2. ECG Signal Segmentation
2.3. Denoising of Electrocardiogram Signals
3. Information Feature Extraction Strategy Based on Relative Position Matrix
3.1. Relative Position Matrix Algorithm
- 1.
- To obtain a standard normal distribution Z for the ECG signal. The z-score normalization can be performed as:
- 2.
- Calculate the relative position between two time steps and transform the pre-processed ECG signal X into a two-dimensional matrix M. Each value at time step i serves as the reference point for each row of M. The transformation equation is formulated as follows:
- 3.
- The final gray-level matrix F is obtained by applying minimum–maximum normalization below:
3.2. Conversion of Relative Position Matrix ECG Image
4. Design of Gam-Resnet18 Network Model for Relative Position Matrix Recognition
5. Implementation of Arrhythmia Classification Based on Gam-Resnet18 Network
5.1. Gam-Resnet18 Network Training
5.2. Metrics Evaluation
5.3. Identification Results
6. Comparison of Images for Classification Modeling with Gam-Resnet18 Network
6.1. Transformation of ECG Signals
6.2. Modeling Results
6.3. Comparison with the Original ECG Signal
6.4. Comparison with Reported Results
6.5. Discussions
7. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Data Availability Statement
Conflicts of Interest
References
- Qin, Q.; Li, J.; Zhang, L.; Yue, Y.; Liu, C. Combining Low-dimensional wavelet features and support vector machine for arrhythmia beat classification. Sci. Rep. 2017, 7, 6067. [Google Scholar] [CrossRef] [PubMed]
- Yang, S. Design and Implementation of an Arrhythmia Diagnosis Algorithm Based on Deep Learning. Ph.D. Dissertation, University of Chinese Academy of Sciences, Beijing, China, 2019. [Google Scholar]
- Zhu, W.; Chen, X.; Wang, Y.; Wang, L. Arrhythmia Recognition and Classification Using ECG Morphology and Segment Feature Analysis. IEEE/ACM Trans. Comput. Biol. Bioinform. 2019, 16, 131–138. [Google Scholar] [CrossRef]
- Elhaj, F.A.; Salim, N.; Harris, A.R.; Swee, T.T.; Ahmed, T. Arrhythmia recognition and classification using combined linear and nonlinear features of ECG signals. Comput. Methods Programs Biomed. 2016, 127, 52–63. [Google Scholar] [CrossRef] [PubMed]
- Li, H.; Yuan, D.; Ma, X.; Cui, D.; Cao, L. Genetic algorithm for the optimization of features and neural networks in ECG signals classification. Sci. Rep. 2017, 7, 41011. [Google Scholar] [CrossRef] [PubMed]
- Zarei, A.; Asl, B.M. Automatic Detection of Obstructive Sleep Apnea Using Wavelet Transform and Entropy-Based Features from Single-Lead ECG Signal. IEEE J. Biomed. Health Inform. 2019, 23, 1011–1021. [Google Scholar] [CrossRef] [PubMed]
- Lu, W.; Shuai, J.; Gu, S.; Xue, J. Method to Annotate Arrhythmias by Deep Network. In Proceedings of the 2018 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData), Halifax, NS, Canada, 30 July–3 August 2018; pp. 1848–1851. [Google Scholar]
- Zhang, G. Research and Development of a Robust Classification Algorithm and Platform for Cardiovascular Diseases. Ph.D. Dissertation, Jilin University, Jilin, China, 2021. [Google Scholar]
- Liu, Z. ECG Data Recognition Algorithm Based on Variable-Scale Fusion Network Model. J. Biomed. Eng. 2022, 39, 570–578. [Google Scholar]
- Zhai, X.; Tin, C. Automated ECG Classification Using Dual Heartbeat Coupling Based on Convolutional Neural Network. IEEE Access 2018, 6, 27465–27472. [Google Scholar] [CrossRef]
- Zhang, Y.; Zhao, Z.; Deng, Y.; Zhang, X.; Zhang, Y. Human identification driven by deep CNN and transfer learning based on multiview feature representations of ECG. Biomed. Signal Process. 2021, 68, 102689. [Google Scholar] [CrossRef]
- Zhang, G. Arrhythmia Classification System Based on ResGC-Net. Ph.D. Dissertation, Jilin University, Jilin, China, 2022. [Google Scholar]
- Zuo, L. Classification of Arrhythmia Based on Semi-supervised Convolutional Neural Network. Laser J. 2021, 42, 219–225. [Google Scholar]
- MIT-BIH Arrhythmia Database v1.0.0. Available online: https://physionet.org/ (accessed on 5 September 2022).
- Yang, S.; Gui, B.; Chen, S. Arrhythmia Detection Based on Wavelet Decomposition and 1D-GoogLeNet. J. Electron. Inf. Technol. 2021, 43, 3018–3027. [Google Scholar]
- Wei, G. Denoising Method of ECG Signals Based on Wavelet Transform. Chin. Med. Equip. J. 2010, 7, 9–11. [Google Scholar]
- Chen, W.; Shi, K. A deep learning framework for time series classification using Relative Position Matrix and Convolutional Neural Network. Neurocomputing 2019, 359, 384–394. [Google Scholar] [CrossRef]
- He, K.; Zhang, X.; Ren, S.; Sun, J. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 27–30 June 2016; pp. 770–778. [Google Scholar]
- Liu, Y.; Shao, Z.; Hoffmann, N. Global Attention Mechanism: Retain Information to Enhance Channel-Spatial Interactions. arXiv 2021, arXiv:2112.05561. [Google Scholar]
- Wang, Z.; Oates, T. Imaging Time-Series to Improve Classification and Imputation; AAAI Press: Washington, DC, USA, 2015. [Google Scholar]
- Eckmann, J.P.; Kamphorst, S.O.; Ruelle, D. Recurrence plots of dynamical systems. World Sci. Ser. Nonlinear Sci. Ser. A 1995, 16, 441–446. [Google Scholar]
- Zha, X.F.; Yang, F.; Wu, Y.N.; Liu, Y.; Yuan, S.F. ECG classification based on transfer learning and deep convolution neural network. Chin. J. Med. Phys. 2018, 35, 1307–1312. [Google Scholar]
- Wang, J.; Shi, M.; Zhang, X. Research on Classification of Arrhythmia Based on EMD and ApEn Feature Extraction. J. Instrum. Meas. 2016, 37, 168–173. [Google Scholar]
- Oh, S.L.; Ng, E.Y.; San Tan, R.; Acharya, U.R. Automated diagnosis of arrhythmia using combination of CNN and LSTM techniques with variable length heart beats. Comput. Biol. Med. 2018, 102, 278–287. [Google Scholar] [CrossRef] [PubMed]
- Kandala, N.V.; Dhuli, R.R. Classification of ECG heartbeats using nonlinear decomposition methods and support vector machine. Comput. Biol. Med. 2017, 87, 271–284. [Google Scholar]
Heart Rate Types | A | V | L | R | N |
---|---|---|---|---|---|
Number | 1950 | 6974 | 6578 | 4967 | 71,723 |
Heart Rate Types | A | V | L | R | N |
---|---|---|---|---|---|
Number | 1950 | 6974 | 6578 | 4967 | 8996 |
Segment | Type | ACC | PPV | SP | SE | F1 Score | Average PPV | Average SE | Average SP |
---|---|---|---|---|---|---|---|---|---|
Mixed heartbeat | A | 99.30% | 95.7% | 99.7% | 96.9% | 96.3% | 98.76% | 99.84% | 98.90% |
L | 99.3% | 99.8% | 99.7% | 99.5% | |||||
N | 99.9% | 100.0% | 100.0% | 99.9% | |||||
R | 99.7% | 99.9% | 99.0% | 99.3% | |||||
V | 99.2% | 99.8% | 98.9% | 99.0% | |||||
Single beat | A | 99.07% | 97.8% | 99.9% | 82.3% | 89.4% | 98.74% | 99.30% | 95.22% |
L | 98.6% | 99.8% | 99.2% | 98.9% | |||||
N | 99.2% | 97.1% | 99.8% | 99.5% | |||||
R | 99.7% | 99.9% | 99.0% | 99.3% | |||||
V | 98.4% | 99.8% | 95.8% | 97.0% |
Method | Type | ACC | PPV | SP | SE | F1 Score | Average PPV | Average SE | Average SP |
---|---|---|---|---|---|---|---|---|---|
RP | A | 99.15% | 94.9% | 99.6% | 96.2% | 95.5% | 98.50% | 98.72% | 99.64% |
L | 99.0% | 99.7% | 99.8% | 99.4% | |||||
N | 100% | 100% | 100% | 100% | |||||
R | 99.3% | 99.9% | 99.5% | 99.4% | |||||
V | 99.3% | 99.8% | 98.1% | 98.7% | |||||
GAF | A | 99.28% | 96.2% | 99.7% | 97.4% | 96.8% | 98.80% | 98.89% | 99.82% |
L | 99.0% | 99.7% | 99.5% | 99.2% | |||||
N | 100.0% | 100.0% | 100.0% | 100.0% | |||||
R | 99.5% | 99.9% | 99.4% | 99.4% | |||||
V | 99.3% | 99.8% | 98.6% | 98.9% | |||||
MTF | A | 98.57% | 95.0% | 99.7% | 92.3% | 93.6% | 97.92% | 97.62% | 99.68% |
L | 98.6% | 99.6% | 99.2% | 98.9% | |||||
N | 99.9% | 100.0% | 99.9% | 99.9% | |||||
R | 98.5% | 99.7% | 99.1% | 98.8% | |||||
V | 97.9% | 99.4% | 97.6% | 97.7% |
Type | ACC | PPV | SP | SE | F1 Score | Average PPV | Average SP | Average SE |
---|---|---|---|---|---|---|---|---|
A | 99.27% | 96.2% | 99.7% | 97.7% | 96.9% | 98.76% | 99.80% | 99.02% |
L | 99.5% | 99.8% | 98.9% | 99.2% | ||||
N | 99.9% | 100% | 100% | 100% | ||||
R | 99.3% | 99.8% | 99.6% | 99.4% | ||||
V | 98.9% | 99.7% | 98.9% | 98.9% |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Zhang, M.; Jin, H.; Zheng, B.; Luo, W. Deep Learning Modeling of Cardiac Arrhythmia Classification on Information Feature Fusion Image with Attention Mechanism. Entropy 2023, 25, 1264. https://doi.org/10.3390/e25091264
Zhang M, Jin H, Zheng B, Luo W. Deep Learning Modeling of Cardiac Arrhythmia Classification on Information Feature Fusion Image with Attention Mechanism. Entropy. 2023; 25(9):1264. https://doi.org/10.3390/e25091264
Chicago/Turabian StyleZhang, Mingming, Huiyuan Jin, Bin Zheng, and Wenbo Luo. 2023. "Deep Learning Modeling of Cardiac Arrhythmia Classification on Information Feature Fusion Image with Attention Mechanism" Entropy 25, no. 9: 1264. https://doi.org/10.3390/e25091264