Multi-stream Bi-GRU network to extract a comprehensive feature set for ECG signal classification

Allam, Jaya Prakash; Sahoo, Suraj Prakash; Ari, Samit

doi:10.1016/j.bspc.2024.106097

Cited by 5 publications

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

A hybrid deep learning network for automatic diagnosis of cardiac arrhythmia based on 12-lead ECG

Bai,

Dong,

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Cardiac arrhythmias are the leading cause of death and pose a huge health and economic burden globally. Electrocardiography (ECG) is an effective technique for the diagnosis of cardiovascular diseases because of its noninvasive and cost-effective advantages. However, traditional ECG analysis relies heavily on the clinical experience of physicians, which can be challenging and time-consuming to produce valid diagnostic results. This work proposes a new hybrid deep learning model that combines convolutional neural network (CNN) and bidirectional gated recurrent unit (BiGRU) with multi-head attention (CBGM model). Specifically, the model consists of seven convolutional layers with varying filter sizes (4, 16, 32, and 64) and three pooling layers, respectively, while the BiGRU module includes two layers with 64 units each followed by multi-head attention (8-heads). The combination of CNN and BiGRU effectively captures spatio-temporal features of ECG signals, with multi-head attention comprehensively extracted global correlations among multiple segments of ECG signals. The validation in the MIT-BIH arrhythmia database achieved an accuracy of 99.41%, a precision of 99.15%, a specificity of 99.68%, and an F1-Score of 99.21%, indicating its robust performance across different evaluation metrics. Additionally, the model’s performance was evaluated on the PTB Diagnostic ECG Database, where it achieved an accuracy of 98.82%, demonstrating its generalization capability. Comparative analysis against previous methods revealed that our proposed CBGM model exhibits more higher performance in automatic classification of arrhythmia and can be helpful for assisting clinicians by enabling real-time detection of cardiac arrhythmias during routine ECG screenings.

show abstract

A hybrid deep learning network for automatic diagnosis of cardiac arrhythmia based on 12-lead ECG

Bai,

Dong,

et al. 2024

Sci Rep

View full text Add to dashboard Cite

show abstract

Efficient Detection of Brain Tumor Using Deep Learning with Small Dataset

Kumar,

Das

2024

2024 3rd International Conference on Artificial Intelligence for Internet of Things (AIIoT)

View full text Add to dashboard Cite

Real-Time Cardiac Abnormality Monitoring and Nursing for Patient Using Electrocardiographic Signals

Ao,

Zhai,

Jiang

et al. 2024

Cardiology

View full text Add to dashboard Cite

Introduction Cardiovascular disease care is a critical clinical application that necessitates real-time monitoring models. Previous models required the use of multi-lead signals and could not be customized as needed. Traditional methods relied on manually designed supervised algorithms, based on empirical experience, to identify waveform abnormalities and classify diseases, and were incapable of monitoring and alerting abnormalities in individual waveforms. Methods This research reconstructed the vector model for arbitrary leads using the phase space time delay method, enabling the model to arbitrarily combine signals as needed while possessing adaptive denoising capabilities. After employing automatically constructed machine learning algorithms and designing for rapid convergence, the model can identify abnormalities in individual waveforms and classify diseases, as well as detect and alert on abnormal waveforms. Result Effective noise elimination was achieved, obtaining a higher degree of loss function fitting. Afterwards, the detail differences of the electrocardiogram signal were amplified using a single-lead three-dimensional model. A cropping algorithm was used to remove waveforms severely interfered by external factors. Then, automatic neural network recognition was used. The automatic network generation model was designed effectively for different data types. The accuracy of patient identification is 98.2%, and the accuracy for healthy patients is 99.2%. Conclusion The elastic wavelet neural network can automatically denoise. Through the three-dimensional model, the detailed changes of electrocardiogram signals of different diseases can be observed. The cropping algorithm effectively identified the interfered and destroyed waveforms. The automatic neural network is capable of carrying out disease type classification and patient identity classification.

show abstract

Multi-stream Bi-GRU network to extract a comprehensive feature set for ECG signal classification

Cited by 5 publications

References 36 publications

A hybrid deep learning network for automatic diagnosis of cardiac arrhythmia based on 12-lead ECG

A hybrid deep learning network for automatic diagnosis of cardiac arrhythmia based on 12-lead ECG

Efficient Detection of Brain Tumor Using Deep Learning with Small Dataset

Real-Time Cardiac Abnormality Monitoring and Nursing for Patient Using Electrocardiographic Signals

Contact Info

Product

Resources

About