ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE Block
Electrocardiogram (ECG) is nowadays an important technology to be applied in the clinical diagnosis for the detection of the heart disease. But the large storage and high-burden transmission of the ECG data is a challenge. Therefore, the compressive sensing (CS) is appropriate to deal with those sig...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10253948/ |
| _version_ | 1827801285771919360 |
|---|---|
| author | Jing Hua Jiawen Zou Jue Rao Hua Yin Jie Chen |
| author_facet | Jing Hua Jiawen Zou Jue Rao Hua Yin Jie Chen |
| author_sort | Jing Hua |
| collection | DOAJ |
| description | Electrocardiogram (ECG) is nowadays an important technology to be applied in the clinical diagnosis for the detection of the heart disease. But the large storage and high-burden transmission of the ECG data is a challenge. Therefore, the compressive sensing (CS) is appropriate to deal with those signals for it can compress and sample the signal at the same time. In order to get rid of the constraints in the traditional CS methods, we propose a compressive sensing framework based on multiscale feature fusion and SE block. In the compression process we use sequential convolutional layers instead of the traditional compressive sensing using measurement matrix projection for ECG signals. In the reconstruction process, the multi-scale feature fusion method is first used to fuse multiple feature maps output from the convolution layer to better extract signal features. Subsequently, Squeeze-and-Excitation (SE) block is used to further enhance the feature representation. Finally, sequence modeling of the ECG signal is performed using LSTM to obtain the reconstructed signal. The results show that the proposed method performs well on various datasets and evaluation metrics, in the case of SR = 0.4, the PRD and SNR of the experiments on the MIT-BIH Arrhythmia database are 1.55% and 37.66dB, respectively. The PRD and SNR of the experiments on the Non-Invasive Fetal ECG Arrhythmia Database were 2.48% and 34.57dB, respectively, which were the lowest among all the comparison methods, indicating that the proposed method has good ECG signal processing capability. |
| first_indexed | 2024-03-11T20:22:13Z |
| format | Article |
| id | doaj.art-4dc92de90235440e9db9dd21bdc0554b |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-11T20:22:13Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-4dc92de90235440e9db9dd21bdc0554b2023-10-02T23:01:34ZengIEEEIEEE Access2169-35362023-01-011110435910437210.1109/ACCESS.2023.331648710253948ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE BlockJing Hua0https://orcid.org/0000-0002-4715-3077Jiawen Zou1https://orcid.org/0000-0002-5065-3558Jue Rao2Hua Yin3https://orcid.org/0000-0003-4611-8533Jie Chen4School of Software, Jiangxi Agricultural University, Nanchang, ChinaSchool of Computer and Information Engineering, Jiangxi Agricultural University, Nanchang, ChinaSchool of Computer and Information Engineering, Jiangxi Agricultural University, Nanchang, ChinaSchool of Software, Jiangxi Agricultural University, Nanchang, ChinaSchool of Computer and Information Engineering, Jiangxi Agricultural University, Nanchang, ChinaElectrocardiogram (ECG) is nowadays an important technology to be applied in the clinical diagnosis for the detection of the heart disease. But the large storage and high-burden transmission of the ECG data is a challenge. Therefore, the compressive sensing (CS) is appropriate to deal with those signals for it can compress and sample the signal at the same time. In order to get rid of the constraints in the traditional CS methods, we propose a compressive sensing framework based on multiscale feature fusion and SE block. In the compression process we use sequential convolutional layers instead of the traditional compressive sensing using measurement matrix projection for ECG signals. In the reconstruction process, the multi-scale feature fusion method is first used to fuse multiple feature maps output from the convolution layer to better extract signal features. Subsequently, Squeeze-and-Excitation (SE) block is used to further enhance the feature representation. Finally, sequence modeling of the ECG signal is performed using LSTM to obtain the reconstructed signal. The results show that the proposed method performs well on various datasets and evaluation metrics, in the case of SR = 0.4, the PRD and SNR of the experiments on the MIT-BIH Arrhythmia database are 1.55% and 37.66dB, respectively. The PRD and SNR of the experiments on the Non-Invasive Fetal ECG Arrhythmia Database were 2.48% and 34.57dB, respectively, which were the lowest among all the comparison methods, indicating that the proposed method has good ECG signal processing capability.https://ieeexplore.ieee.org/document/10253948/Deep learningcompressive sensingmulti-scale featureSE blockLSTM |
| spellingShingle | Jing Hua Jiawen Zou Jue Rao Hua Yin Jie Chen ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE Block IEEE Access Deep learning compressive sensing multi-scale feature SE block LSTM |
| title | ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE Block |
| title_full | ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE Block |
| title_fullStr | ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE Block |
| title_full_unstemmed | ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE Block |
| title_short | ECG Signals Deep Compressive Sensing Framework Based on Multiscale Feature Fusion and SE Block |
| title_sort | ecg signals deep compressive sensing framework based on multiscale feature fusion and se block |
| topic | Deep learning compressive sensing multi-scale feature SE block LSTM |
| url | https://ieeexplore.ieee.org/document/10253948/ |
| work_keys_str_mv | AT jinghua ecgsignalsdeepcompressivesensingframeworkbasedonmultiscalefeaturefusionandseblock AT jiawenzou ecgsignalsdeepcompressivesensingframeworkbasedonmultiscalefeaturefusionandseblock AT juerao ecgsignalsdeepcompressivesensingframeworkbasedonmultiscalefeaturefusionandseblock AT huayin ecgsignalsdeepcompressivesensingframeworkbasedonmultiscalefeaturefusionandseblock AT jiechen ecgsignalsdeepcompressivesensingframeworkbasedonmultiscalefeaturefusionandseblock |