Identification of Atrial Fibrillation from Electrocardiogram Signals Based on Deep Neural Network

2019 ◽  
Vol 9 (4) ◽  
pp. 838-846 ◽  
Author(s):  
Lili Chen ◽  
Ying He
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Deep Neural Network ◽  
Electrocardiogram Signals

Reliable detection of atrial fibrillation with a medical wearable under inpatient conditions (CoMMoD-A-fib)

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
M Jacobsen ◽  
T.A Dembek ◽  
A.P Ziakos ◽  
G Kobbe ◽  
M Kollmann ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Deep Neural Network ◽  
Sampling Rate ◽  
Public Institution ◽  
Funding Source ◽  
Upper Arm ◽  
Recording Time ◽  
Reduced Ejection Fraction ◽  
Non Invasive

Abstract Background Atrial fibrillation (A-fib) is the most common arrhythmia; however, detection of A-fib is a challenge due to irregular occurrence. Purpose Evaluating feasibility and performance of a non-invasive medical wearable for detection of A-fib. Methods In the CoMMoD-A-fib trial admitted patients with a high risk for A-fib carried the wearable and an ECG Holter (control) in parallel over a period of 24 hours under not physically restricted conditions. The wearable with a tight-fit upper arm band employs a photoplethysmography (PPG) technology enabling a high sampling rate. Different algorithms (including a deep neural network) were applied to 5 min PPG datasets for detection of A-fib. Proportion of monitoring time automatically interpretable by algorithms (= interpretable time) was analyzed for influencing factors. Results In 102 inpatients (age 71.0±11.9 years; 52% male) 2306 hours of parallel recording time could be obtained; 1781 hours (77.2%) of these were automatically interpretable by an algorithm analyzing PPG derived intervals. Detection of A-Fib was possible with a sensitivity of 92.7% and specificity of 92.4% (AUC 0.96). Also during physical activity, detection of A-fib was sufficiently possible (sensitivity 90.1% and specificity 91.2%). Usage of the deep neural network improved detection of A-fib further (sensitivity 95.4% and specificity 96.2%). A higher prevalence of heart failure with reduced ejection fraction was observed in patients with a low interpretable time (p=0.080). Conclusion Detection of A-fib by means of an upper arm non-invasive medical wearable with a high resolution is reliably possible under inpatient conditions. Funding Acknowledgement Type of funding source: Public Institution(s). Main funding source(s): Internal grant program (PhD and Dr. rer. nat. Program Biomedicine) of the Faculty of Health at Witten/Herdecke University, Germany. HELIOS Kliniken GmbH (Grant-ID 047476), Germany

scholarly journals Atrial Fibrillation Detection by the Combination of Recurrence Complex Network and Convolution Neural Network

2019 ◽  
Vol 2019 ◽  
pp. 1-9 ◽  
Author(s):  
Xiaoling Wei ◽  
Jimin Li ◽  
Chenghao Zhang ◽  
Ming Liu ◽  
Peng Xiong ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Complex Network ◽  
Individual Variation ◽  
Deep Neural Network ◽  
Detection Algorithm ◽  
Convolution Neural Network ◽  
Electrocardiogram Signal ◽  
Sensitivity Specificity ◽  
Atrial Fibrillation Detection

In this paper, R wave peak interval independent atrial fibrillation detection algorithm is proposed based on the analysis of the synchronization feature of the electrocardiogram signal by a deep neural network. Firstly, the synchronization feature of each heartbeat of the electrocardiogram signal is constructed by a Recurrence Complex Network. Then, a convolution neural network is used to detect atrial fibrillation by analyzing the eigenvalues of the Recurrence Complex Network. Finally, a voting algorithm is developed to improve the performance of the beat-wise atrial fibrillation detection. The MIT-BIH atrial fibrillation database is used to evaluate the performance of the proposed method. Experimental results show that the sensitivity, specificity, and accuracy of the algorithm can achieve 94.28%, 94.91%, and 94.59%, respectively. Remarkably, the proposed method was more effective than the traditional algorithms to the problem of individual variation in the atrial fibrillation detection.

Detection of Atrial Fibrillation from Single Lead ECG Signal Using Multirate Cosine Filter Bank and Deep Neural Network

2020 ◽  
Vol 44 (6) ◽  
Author(s):  
S. K. Ghosh ◽  
R. K. Tripathy ◽  
Mario R. A. Paternina ◽  
Juan J. Arrieta ◽  
Alejandro Zamora-Mendez ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Filter Bank ◽  
Deep Neural Network ◽  
Ecg Signal

Accurate detection of atrial fibrillation from 12-lead ECG using deep neural network

2020 ◽  
Vol 116 ◽  
pp. 103378 ◽  
Author(s):  
Wenjuan Cai ◽  
Yundai Chen ◽  
Jun Guo ◽  
Baoshi Han ◽  
Yajun Shi ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Deep Neural Network ◽  
Accurate Detection

scholarly journals ATRIAL FIBRILLATION DETECTION ON ELECTROCARDIOGRAMS WITH CONVOLUTIONAL NEURAL NETWORKS

2019 ◽  
Vol 9 (4) ◽  
pp. 69-73
Author(s):  
Viktor Kifer ◽  
Natalia Zagorodna ◽  
Olena Hevko
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Sinus Rhythm ◽  
Deep Neural Network ◽  
Normal Sinus Rhythm ◽  
Normal Sinus ◽  
Noisy Signals ◽  
Network Usage ◽  
Model Training

In this paper, we present our research which confirms the suitability of the convolutional neural network usage for the classification of single-lead ECG recordings. The proposed method was designed for classifying normal sinus rhythm, atrial fibrillation (AF), non-AF related other abnormal heart rhythms and noisy signals. The method combines manually selected features with the features learned by the deep neural network. The Physionet Challenge 2017 dataset of over 8500 ECG recordings was used for the model training and validation. The trained model reaches an average F1-score 0.71 in classifying normal sinus rhythm, AF and other rhythms respectively.

scholarly journals Unravelling the black box of machine learning for atrial fibrillation forecast: role of heart rate variability and of premature beats

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
J.M Gregoire ◽  
C Gilon ◽  
S Carlier ◽  
H Bersini
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Atrial Fibrillation ◽  
Heart Rate ◽  
Heart Rate Variability ◽  
Deep Neural Network ◽  
Black Box ◽  
Rr Intervals ◽  
Network Layers ◽  
Premature Beats

Abstract Background The identification of patients still in sinus rhythm who will present one month later an atrial fibrillation episode is possible using machine learning (ML) techniques. However, these new ML algorithms do not provide any relevant information about the underlying pathophysiology. Purpose To compare the predictive performance for forecasting AF between a machine learning algorithm and other parameters whose pathophysiological mechanisms are known to play a role in the triggering of arrhythmias (i.e. the count of premature beats (PB) and heart rate variability (HRV) parameters) Material and methods We conducted a retrospective study from an outpatient clinic. 10484 Holter ECG recordings were screened. 250 analysable AF onsets were labelled. We developed a deep neural network model composed of convolutional neural network layers and bidirectional gated recurrent units as recurrent neural network layers that was trained for the forecast of paroxysmal AF episodes, using RR intervals variations. This model works like a black box. For comparison purposes, we used a “random forest” (RF) model of ML to obtain forecast results using HRV parameters with and without PB. This model allows the evaluation of the relevance of HRV parameters and of PB used for the forecast. We calculated the area under the curve of the receiving operating characteristic curve for the different time windows counted in RR intervals before the AF onset. Results As shown in the table, the forecasting value of the deep neural network model (ML) was not superior to the random forest algorithm. Prediction value of both decreased when analyzing the RR intervals further away from the onset of AF Conclusions These results suggest that HRV plays a predominant role in triggering AF episodes and that premature beats could add minor information. Moreover, the closer the window from AF onset, the better the accuracy, regardless of the method used. Such detection algorithms once implemented in pacemakers, might prove useful to prevent AF onset by changing pacing sequence while patients would still be in sinus rhythm, however this remains to be demonstrated Funding Acknowledgement Type of funding source: None

Sign in / Sign up

Export Citation Format

Share Document