A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification

Background: Cardiovascular diseases (CVD) are the leading cause of death globally. Electrocardiogram (ECG) analysis can provide thoroughly assessment for different CVDs efficiently. We propose a multi-task group bidirectional long short-term memory (MTGBi-LSTM) framework to intelligent recognize mul...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Qiu-Jie Lv, Hsin-Yi Chen, Wei-Bin Zhong, Ying-Ying Wang, Jing-Yan Song, Sai-Di Guo, Lian-Xin Qi, Calvin Yu-Chian Chen
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2020
Materias:	ECG bidirectional long short-term memory network attention mechanism multi-task learning Computer applications to medicine. Medical informatics R858-859.7 Medical technology R855-855.5
Acceso en línea:	https://doaj.org/article/94f2b2040d264ed7ac2bb0956e9dfd1c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:94f2b2040d264ed7ac2bb0956e9dfd1c
record_format	dspace
spelling	oai:doaj.org-article:94f2b2040d264ed7ac2bb0956e9dfd1c2021-11-19T00:00:24ZA Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification2168-237210.1109/JTEHM.2019.2952610https://doaj.org/article/94f2b2040d264ed7ac2bb0956e9dfd1c2020-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/8896884/https://doaj.org/toc/2168-2372Background: Cardiovascular diseases (CVD) are the leading cause of death globally. Electrocardiogram (ECG) analysis can provide thoroughly assessment for different CVDs efficiently. We propose a multi-task group bidirectional long short-term memory (MTGBi-LSTM) framework to intelligent recognize multiple CVDs based on multi-lead ECG signals. Methods: This model employs a Group Bi-LSTM (GBi-LSTM) and Residual Group Convolutional Neural Network (Res-GCNN) to learn the dual feature representation of ECG space and time series. GBi-LSTM is divided into Global Bi-LSTM and Intra-Group Bi-LSTM, which can learn the features of each ECG lead and the relationship between leads. Then, through attention mechanism, the different lead information of ECG is integrated to make the model to possess the powerful feature discriminability. Through multi-task learning, the model can fully mine the association information between diseases and obtain more accurate diagnostic results. In addition, we propose a dynamic weighted loss function to better quantify the loss to overcome the imbalance between classes. Results: Based on more than 170,000 clinical 12-lead ECG analysis, the MTGBi-LSTM method achieved accuracy, precision, recall and F1 of 88.86%, 90.67%, 94.19% and 92.39%, respectively. The experimental results show that the proposed MTGBi-LSTM method can reliably realize ECG analysis and provide an effective tool for computer-aided diagnosis of CVD.Qiu-Jie LvHsin-Yi ChenWei-Bin ZhongYing-Ying WangJing-Yan SongSai-Di GuoLian-Xin QiCalvin Yu-Chian ChenIEEEarticleECGbidirectional long short-term memory networkattention mechanismmulti-task learningComputer applications to medicine. Medical informaticsR858-859.7Medical technologyR855-855.5ENIEEE Journal of Translational Engineering in Health and Medicine, Vol 8, Pp 1-11 (2020)
institution	DOAJ
collection	DOAJ
language	EN
topic	ECG bidirectional long short-term memory network attention mechanism multi-task learning Computer applications to medicine. Medical informatics R858-859.7 Medical technology R855-855.5
spellingShingle	ECG bidirectional long short-term memory network attention mechanism multi-task learning Computer applications to medicine. Medical informatics R858-859.7 Medical technology R855-855.5 Qiu-Jie Lv Hsin-Yi Chen Wei-Bin Zhong Ying-Ying Wang Jing-Yan Song Sai-Di Guo Lian-Xin Qi Calvin Yu-Chian Chen A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification
description	Background: Cardiovascular diseases (CVD) are the leading cause of death globally. Electrocardiogram (ECG) analysis can provide thoroughly assessment for different CVDs efficiently. We propose a multi-task group bidirectional long short-term memory (MTGBi-LSTM) framework to intelligent recognize multiple CVDs based on multi-lead ECG signals. Methods: This model employs a Group Bi-LSTM (GBi-LSTM) and Residual Group Convolutional Neural Network (Res-GCNN) to learn the dual feature representation of ECG space and time series. GBi-LSTM is divided into Global Bi-LSTM and Intra-Group Bi-LSTM, which can learn the features of each ECG lead and the relationship between leads. Then, through attention mechanism, the different lead information of ECG is integrated to make the model to possess the powerful feature discriminability. Through multi-task learning, the model can fully mine the association information between diseases and obtain more accurate diagnostic results. In addition, we propose a dynamic weighted loss function to better quantify the loss to overcome the imbalance between classes. Results: Based on more than 170,000 clinical 12-lead ECG analysis, the MTGBi-LSTM method achieved accuracy, precision, recall and F1 of 88.86%, 90.67%, 94.19% and 92.39%, respectively. The experimental results show that the proposed MTGBi-LSTM method can reliably realize ECG analysis and provide an effective tool for computer-aided diagnosis of CVD.
format	article
author	Qiu-Jie Lv Hsin-Yi Chen Wei-Bin Zhong Ying-Ying Wang Jing-Yan Song Sai-Di Guo Lian-Xin Qi Calvin Yu-Chian Chen
author_facet	Qiu-Jie Lv Hsin-Yi Chen Wei-Bin Zhong Ying-Ying Wang Jing-Yan Song Sai-Di Guo Lian-Xin Qi Calvin Yu-Chian Chen
author_sort	Qiu-Jie Lv
title	A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification
title_short	A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification
title_full	A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification
title_fullStr	A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification
title_full_unstemmed	A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification
title_sort	multi-task group bi-lstm networks application on electrocardiogram classification
publisher	IEEE
publishDate	2020
url	https://doaj.org/article/94f2b2040d264ed7ac2bb0956e9dfd1c
work_keys_str_mv	AT qiujielv amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT hsinyichen amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT weibinzhong amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT yingyingwang amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT jingyansong amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT saidiguo amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT lianxinqi amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT calvinyuchianchen amultitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT qiujielv multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT hsinyichen multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT weibinzhong multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT yingyingwang multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT jingyansong multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT saidiguo multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT lianxinqi multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification AT calvinyuchianchen multitaskgroupbilstmnetworksapplicationonelectrocardiogramclassification
_version_	1718420680494546944

A Multi-Task Group Bi-LSTM Networks Application on Electrocardiogram Classification

Ejemplares similares