TY - GEN
T1 - Deep Learning-Based Approach for Atrial Fibrillation Detection
AU - Khriji, Lazhar
AU - Fradi, Marwa
AU - Machhout, Mohsen
AU - Hossen, Abdulnasir
N1 - Funding Information:
The authors would like to thank OMANTEL and Sultan Qaboos University for their financial support, grant number “EG/SQU-OT/18/01”.
Publisher Copyright:
© 2020, The Author(s).
PY - 2020
Y1 - 2020
N2 - Atrial Fibrillation (AF) is a health-threatening condition, which is a violation of the heart rhythm that can lead to heart-related complications. Remarkable interest has been given to ECG signals analysis for AF detection in an early stage. In this context, we propose an artificial neural network ANN application to classify ECG signals into three classes, the first presents Normal Sinus Rhythm NSR, the second depicts abnormal signal with Atrial Fibrillation (AF) and the third shows noisy ECG signals. Accordingly, we achieve 93.1% accuracy classification results, 95.1% of sensitivity, 90.5% of specificity and 98%. Furthermore, we yield a value of zero error and a low value of cross entropy, which prove the robustness of the proposed ANN model architecture. Thus, we outperform the state of the art by achieving high accuracy classification without pre-processing step and without high level of feature extraction, and then we enable clinicians to determine automatically the class of each patient ECG signal.
AB - Atrial Fibrillation (AF) is a health-threatening condition, which is a violation of the heart rhythm that can lead to heart-related complications. Remarkable interest has been given to ECG signals analysis for AF detection in an early stage. In this context, we propose an artificial neural network ANN application to classify ECG signals into three classes, the first presents Normal Sinus Rhythm NSR, the second depicts abnormal signal with Atrial Fibrillation (AF) and the third shows noisy ECG signals. Accordingly, we achieve 93.1% accuracy classification results, 95.1% of sensitivity, 90.5% of specificity and 98%. Furthermore, we yield a value of zero error and a low value of cross entropy, which prove the robustness of the proposed ANN model architecture. Thus, we outperform the state of the art by achieving high accuracy classification without pre-processing step and without high level of feature extraction, and then we enable clinicians to determine automatically the class of each patient ECG signal.
KW - AF detection
KW - ANN
KW - Confusion matrix
KW - ECG-classification
KW - Histogram error
KW - ROC
UR - http://www.scopus.com/inward/record.url?scp=85087778670&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85087778670&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-51517-1_9
DO - 10.1007/978-3-030-51517-1_9
M3 - Conference contribution
AN - SCOPUS:85087778670
SN - 9783030515164
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 100
EP - 113
BT - The Impact of Digital Technologies on Public Health in Developed and Developing Countries - 18th International Conference, ICOST 2020, Proceedings
A2 - Jmaiel, Mohamed
A2 - Aloulou, Hamdi
A2 - Mokhtari, Mounir
A2 - Abdulrazak, Bessam
A2 - Kallel, Slim
PB - Springer
T2 - 18th International Conference on Smart Homes and Health Telematics, ICOST 2020
Y2 - 24 June 2020 through 26 June 2020
ER -