Classification of pathological ECG beats based on wireless body sensor networks and fractional Fourier transform and convolutional neural network.
Saved in:
| Title: | Classification of pathological ECG beats based on wireless body sensor networks and fractional Fourier transform and convolutional neural network. |
|---|---|
| Authors: | Chaabane, Mohamed1 (AUTHOR), Chehri, Abdellah2 (AUTHOR), Saadane, Rachid3 (AUTHOR), Jeon, Gwanggil4 (AUTHOR) gjeon@inu.ac.kr, El Rharras, Abdessamad1 (AUTHOR) |
| Source: | Wireless Networks (10220038). Nov2024, Vol. 30 Issue 8, p7059-7073. 15p. |
| Subjects: | Convolutional neural networks, Body sensor networks, Body area networks, Wireless sensor networks, Artificial intelligence |
| Abstract: | The heart that beats with an irregular rhythm, sometimes too fast, quickly gets tired. Even though it is the most common cardiac abnormality, atrial fibrillation remains very complex to diagnose. The electrocardiogram (ECG) contains much information about the functioning and possible pathologies of the heart. Artificial intelligence is reshaping healthcare and can be used to diagnose health abnormalities or predict events. The paper aims to combine advanced ECG signal processing with a deep learning technique. After using several filtering algorithms, we divide the ECG signal and then convert it into the frequency domain, precisely the fractional Fourier transform, then into small equal parts so that each contains the typical features of the signal. We collect them in the form of simple arrays. The next step is to convert these arrays into images, which involves transforming each one-dimensional ECG beat into a two-dimensional image. We have relied on this MIT-BIH database research. In the classification phase, 80% was used as training data and 20% for validation and testing. In the end, we classified five categories of heart disease. The proposed technique shows promising results, compared to other similar works, with an accuracy rate of 99.34%. [ABSTRACT FROM AUTHOR] |
| Copyright of Wireless Networks (10220038) is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | The heart that beats with an irregular rhythm, sometimes too fast, quickly gets tired. Even though it is the most common cardiac abnormality, atrial fibrillation remains very complex to diagnose. The electrocardiogram (ECG) contains much information about the functioning and possible pathologies of the heart. Artificial intelligence is reshaping healthcare and can be used to diagnose health abnormalities or predict events. The paper aims to combine advanced ECG signal processing with a deep learning technique. After using several filtering algorithms, we divide the ECG signal and then convert it into the frequency domain, precisely the fractional Fourier transform, then into small equal parts so that each contains the typical features of the signal. We collect them in the form of simple arrays. The next step is to convert these arrays into images, which involves transforming each one-dimensional ECG beat into a two-dimensional image. We have relied on this MIT-BIH database research. In the classification phase, 80% was used as training data and 20% for validation and testing. In the end, we classified five categories of heart disease. The proposed technique shows promising results, compared to other similar works, with an accuracy rate of 99.34%. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 10220038 |
| DOI: | 10.1007/s11276-023-03566-4 |
