DAE-ConvBiLSTM: End-to-end learning single-lead electrocardiogram signal for heart abnormalities detection

Abstract

Background The electrocardiogram (ECG) is a widely used diagnostic that observes the heart activities of patients to ascertain a heart abnormality diagnosis. The artifacts or noises are primarily associated with the problem of ECG signal processing. Conventional denoising techniques have been proposed in previous literature; however, some lacks, such as the determination of suitable wavelet basis function and threshold, can be a time-consuming process. This paper presents end-to-end learning using a denoising auto-encoder (DAE) for denoising algorithms and convolutional-bidirectional long short-term memory (ConvBiLSTM) for ECG delineation to classify ECG waveforms in terms of the PQRST-wave and isoelectric lines. The denoising reconstruction using unsupervised learning based on the encoder-decoder process can be proposed to improve the drawbacks. First, The ECG signals are reduced to a low-dimensional vector in the encoder. Second, the decoder reconstructed the signals. The last, the reconstructed signals of ECG can be processed to ConvBiLSTM. The proposed architecture of DAE-ConvBiLSTM is the end-to-end diagnosis of heart abnormality detection. Results As a result, the performance of DAE-ConvBiLSTM has obtained an average of above 98.59% accuracy, sensitivity, specificity, precision, and F1 score from the existing studies. The DAE-ConvBiLSTM has also experimented with detecting T-wave (due to ventricular repolarisation) morphology abnormalities. Conclusion The development architecture for detecting heart abnormalities using an unsupervised learning DAE and supervised learning ConvBiLSTM can be proposed for an end-to-end learning algorithm. In the future, the precise accuracy of the ECG main waveform will affect heart abnormalities detection in clinical practice.

Robust electrocardiogram delineation model for automatic morphological abnormality interpretation

“… To test our delineation model, we have explored QT Database (QTDB) as testing set (unseen) to provide an unbiased evaluation of a best model fit on the training dataset 22 . QTDB has commonly experimented for ECG delineation task in several researches 6 , 10 , 18 , 19 . QTDB has been used due to it was provided the beats that manually annotated by cardiologists.…”

“…ECG delineation is a crucial step in processing ECG signals and helps to identify the critical points that indicate the interval and amplitude locations in each wave morphology 6 . There are two main ECG delineation methods: digital signal processing methods 7 – 9 and intelligent processing methods 3 , 10 – 19 . Many researchers have performed considerable research in both types of ECG delineation 3 , 7 – 19 .…”

“…There are two main ECG delineation methods: digital signal processing methods 7 – 9 and intelligent processing methods 3 , 10 – 19 . Many researchers have performed considerable research in both types of ECG delineation 3 , 7 – 19 . However, several deficiencies remain regarding the delineation of ECG signal methods: (1) some existing methods require prespecified thresholds or other assumptions 8 , (2) a limited number of studies have investigated the role of medical knowledge in delineation models as a decision 20 , and (3) the effects of noise have not been closely examined 18 .…”

Empowering AI-Diagnosis: Deep Learning Abilities for Accurate Atrial Fibrillation Classification

“…To generalize the proposed experimental model, we compared the effectiveness of DWT and DAE for ECG denoising. In our previous works [29,30], we explored the use of DAE for removing noise and artifacts from ECG signals. Denoising algorithms based on DL have been explored for performing ECG signal denoising [29][30][31][32][33][34].…”

“…In our previous works [29,30], we explored the use of DAE for removing noise and artifacts from ECG signals. Denoising algorithms based on DL have been explored for performing ECG signal denoising [29][30][31][32][33][34]. DAE learns the parameters for different noisy conditions, which consist of encoding (lowerdimensional representation) and decoding layers (feature extraction).…”