Cardiovascular diseases (CVDs) are the main cause of death worldwide, with arrhythmias posing significant risks. The interpretation of ECG signals requires professional expertise and time, which has limited its accessibility to the general public. This research introduces a novel hybrid deep learning technique for the automatic detection of key features (QRS complex, P and T waves) in electrocardiogram (ECG) signals. The method integrates two Convolutional Neural Networks (CNNs) and two Bidirectional Long Short-Term Memory Networks (BiLSTMs) to form a 2D-CNN-BiLSTM model. The model is trained and evaluated using the "QT Database", a publicly available dataset. ECG signals are first downsampled and then pre-processed using a local linear regression technique to remove the baseline, and the Continuous Wavelet Transform (CWT) to convert 1D ECG signals into 2D images. The model demonstrates superior classification performance in detecting P waves, QRS complexes, and T waves, compared to existing methods. The suggested approach shows promise for improving the precision and effectiveness of ECG analysis and advancing automated cardiac diagnostic techniques.
Enhancing ECG Analysis with a Hybrid Deep Learning Approach: Automatic Detection of Significant Features / Delrio, F.; De Vitis, V.; Caligari, S.; Randazzo, V.; Cirrincione, G.; Pasero, E. (SMART INNOVATION, SYSTEMS AND TECHNOLOGIES). - In: Smart Innovation, Systems and Technologies / Esposito A., Faundez-Zanuy M., Morabito F. C., Pasero E., Cordasco G.. - STAMPA. - Singapore : Springer Nature, 2025. - ISBN 9789819609932. - pp. 23-33 [10.1007/978-981-96-0994-9_3]
Enhancing ECG Analysis with a Hybrid Deep Learning Approach: Automatic Detection of Significant Features
Delrio F.;De Vitis V.;Caligari S.;Randazzo V.;Pasero E.
2025
Abstract
Cardiovascular diseases (CVDs) are the main cause of death worldwide, with arrhythmias posing significant risks. The interpretation of ECG signals requires professional expertise and time, which has limited its accessibility to the general public. This research introduces a novel hybrid deep learning technique for the automatic detection of key features (QRS complex, P and T waves) in electrocardiogram (ECG) signals. The method integrates two Convolutional Neural Networks (CNNs) and two Bidirectional Long Short-Term Memory Networks (BiLSTMs) to form a 2D-CNN-BiLSTM model. The model is trained and evaluated using the "QT Database", a publicly available dataset. ECG signals are first downsampled and then pre-processed using a local linear regression technique to remove the baseline, and the Continuous Wavelet Transform (CWT) to convert 1D ECG signals into 2D images. The model demonstrates superior classification performance in detecting P waves, QRS complexes, and T waves, compared to existing methods. The suggested approach shows promise for improving the precision and effectiveness of ECG analysis and advancing automated cardiac diagnostic techniques.File | Dimensione | Formato | |
---|---|---|---|
2025_bookExtracted_WIRN2023_ECG features extraction_Tesi_DeVitis.pdf
accesso riservato
Tipologia:
2a Post-print versione editoriale / Version of Record
Licenza:
Non Pubblico - Accesso privato/ristretto
Dimensione
844.46 kB
Formato
Adobe PDF
|
844.46 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
WIRN2023_paper_20_postPrint.pdf
embargo fino al 25/07/2026
Tipologia:
2. Post-print / Author's Accepted Manuscript
Licenza:
Pubblico - Tutti i diritti riservati
Dimensione
938.06 kB
Formato
Adobe PDF
|
938.06 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11583/3002108