Research Article
Prediction of Ventricular Tachyarrhythmia Using Deep Learning
@INPROCEEDINGS{10.1007/978-3-031-32029-3_5, author={Dalila Barbosa and E. J. Solteiro Pires and Argentina Leite and P. B. de Moura Oliveira}, title={Prediction of Ventricular Tachyarrhythmia Using Deep Learning}, proceedings={Wireless Mobile Communication and Healthcare. 11th EAI International Conference, MobiHealth 2022, Virtual Event, November 30 -- December 2, 2022, Proceedings}, proceedings_a={MOBIHEALTH}, year={2023}, month={5}, keywords={Ventricular Tachyarrhythmia Deep learning Long Short Term Memory}, doi={10.1007/978-3-031-32029-3_5} }- Dalila Barbosa
E. J. Solteiro Pires
Argentina Leite
P. B. de Moura Oliveira
Year: 2023
Prediction of Ventricular Tachyarrhythmia Using Deep Learning
MOBIHEALTH
Springer
DOI: 10.1007/978-3-031-32029-3_5
Abstract
Ventricular tachyarrhythmia (VTA), mainly ventricular tachycardia (VT) and ventricular fibrillation (VF) are the major causes of sudden cardiac death in the world. This work uses deep learning, more precisely, LSTM and biLSTM networks to predict VTA events. The Spontaneous Ventricular Tachyarrhythmia Database from PhysioNET was chosen, which contains 78 patients, 135 VTA signals, and 135 control rhythms. After the pre-processing of these signals and feature extraction, the classifiers were able to predict whether a patient was going to suffer a VTA event or not. A better result using a biLSTM was obtained, with a 5-fold-cross-validation, reaching an accuracy of 96.30%, 94.07% of precision, 98.45% of sensibility, and 96.17% of F1-Score.
