Towards Early Detection and Burden Estimation of Atrial Fibrillation in an Ambulatory Free-living Environment

2021 ◽  
Vol 5 (2) ◽  
pp. 1-19
Author(s):  
Hanbin Zhang ◽  
Li Zhu ◽  
Viswam Nathan ◽  
Jilong Kuang ◽  
Jacob Kim ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Early Detection ◽  
Daily Living ◽  
Research Study ◽  
Heart Rhythm ◽  
Living Environment ◽  
User Engagement ◽  
Free Living ◽  
Time Frequency

Early detection and accurate burden estimation of atrial fibrillation (AFib) can provide the foundation for effective physician treatment. New approaches to accomplish this have attracted tremendous attention in recent years. In this paper, we develop a novel passive smartwatch-based system to detect AFib episodes and estimate the AFib burden in an ambulatory free-living environment without user engagement. Our system leverages a built-in PPG sensor to collect heart rhythm without user engagement. Then, a data preprocessor module includes time-frequency (TF) analysis to augment features in both the time and frequency domain. Finally, a lightweight multi-view convolutional neural network consisting of 19 layers achieves the AFib detection. To validate our system, we carry out a research study that enrolls 53 participants across three months, where we collect and annotate more than 27,622 hours of data. Our system achieves an average of 91.6% accuracy, 93.0% specificity, and 90.8% sensitivity without dropping any data. Moreover, our system takes 0.51 million parameters and costs 5.18 ms per inference. These results reveal that our proposed system can provide a clinical assessment of AFib in daily living.

scholarly journals K-margin-based Residual-Convolution-Recurrent Neural Network for Atrial Fibrillation Detection

2019 ◽  
Author(s):  
Yuxi Zhou ◽  
Shenda Hong ◽  
Junyuan Shang ◽  
Meng Wu ◽  
Qingyun Wang ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Recurrent Neural Network ◽  
State Of The Art ◽  
Class Imbalance ◽  
Heart Rhythm ◽  
Error Rates ◽  
Detection Task ◽  
Training Data

Atrial Fibrillation (AF) is an abnormal heart rhythm which can trigger cardiac arrest and sudden death. Nevertheless, its interpretation is mostly done by medical experts due to high error rates of computerized interpretation. One study found that only about 66% of AF were correctly recognized from noisy ECGs. This is in part due to insufficient training data, class skewness, as well as semantical ambiguities caused by noisy segments in an ECG record. In this paper, we propose a K-margin-based Residual-Convolution-Recurrent neural network (K-margin-based RCR-net) for AF detection from noisy ECGs. In detail, a skewness-driven dynamic augmentation method is employed to handle the problems of data inadequacy and class imbalance. A novel RCR-net is proposed to automatically extract both long-term rhythm-level and local heartbeat-level characters. Finally, we present a K-margin-based diagnosis model to automatically focus on the most important parts of an ECG record and handle noise by naturally exploiting expected consistency among the segments associated for each record. The experimental results demonstrate that the proposed method with 0.8125 F1NAOP score outperforms all state-of-the-art deep learning methods for AF detection task by 6.8%.

Development of neural network model for signs determination of atrial fibrillation for subsystem of cardiorhythmogram signals processing

2021 ◽  
Author(s):  
Yu.A. Chelebaeva
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Signal Processing ◽  
Network Model ◽  
Neural Network Model ◽  
High Reliability ◽  
Heart Rhythm ◽  
Programmable Logic ◽  
Processing Subsystem

Task of the analysis of a cardio rhythm in real time is detection of early arrhythmias for the purpose of their treatment and prevention of life-endangering arrhythmias. In order to solve the problem of classification of heart rhythm features based on cardiorhythmogram processing, an apparatus of artificial neural networks can be used. One of the most dangerous arrhythmias is atrial fibrillation. Therefore, the development of a neural network model for determining atrial fibrillation features, suitable for implementation on the programmable logic basis, for a subsystem for processing cardiorhythmogram signals is an urgent task. Purpose – development of a neural network model for determining atrial fibrillation features for a signal processing subsystem characterized by high reliability and the implementation possibility on the basis of programmable logic. A neural network model for features determining of atrial fibrillation has been developed, characterized by high reliability and insignificant hardware costs when implemented on field programmable gate arrays (FPGA). Program modeling of neural network model for signs determination of atrial fibrillation is performed. A neural network model for characteristics determining of atrial fibrillation on hardware description language VHDL for use in the signal processing subsystem of a cardiorhythmogram based on FPGA was implemented. The findings suggest that the proposed model can be used in the construction of real-time heart rhythm control systems both for monitoring already diagnosed cardiovascular diseases, especially in intensive care wards, and for the prevention and early diagnosis of arrhythmias in individuals at high myocardial risk.

scholarly journals Smart Care Using a DNN-Based Approach for Activities of Daily Living (ADL) Recognition

2020 ◽  
Vol 11 (1) ◽  
pp. 10
Author(s):  
Muchun Su ◽  
Diana Wahyu Hayati ◽  
Shaowu Tseng ◽  
Jiehhaur Chen ◽  
Hsihsien Wei
Keyword(s):  
Neural Network ◽  
Health Care ◽  
Activities Of Daily Living ◽  
Video Processing ◽  
Daily Living ◽  
Deep Neural Network ◽  
Recognition System ◽  
Automatic Recognition ◽  
Data Filtering ◽  
Accuracy Rate

Health care for independently living elders is more important than ever. Automatic recognition of their Activities of Daily Living (ADL) is the first step to solving the health care issues faced by seniors in an efficient way. The paper describes a Deep Neural Network (DNN)-based recognition system aimed at facilitating smart care, which combines ADL recognition, image/video processing, movement calculation, and DNN. An algorithm is developed for processing skeletal data, filtering noise, and pattern recognition for identification of the 10 most common ADL including standing, bending, squatting, sitting, eating, hand holding, hand raising, sitting plus drinking, standing plus drinking, and falling. The evaluation results show that this DNN-based system is suitable method for dealing with ADL recognition with an accuracy rate of over 95%. The findings support the feasibility of this system that is efficient enough for both practical and academic applications.

scholarly journals Large extracellular vesicles in the left atrial appendage in patients with atrial fibrillation—the missing link?

2021 ◽  
Author(s):  
Andreas Zietzer ◽  
Baravan Al-Kassou ◽  
Paul Jamme ◽  
Verena Rolfes ◽  
Eva Steffen ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
Extracellular Vesicles ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
Thrombus Formation ◽  
Flow Cytometric Analysis ◽  
Heart Rhythm ◽  
Atrial Appendage ◽  
Peripheral Embolization

AbstractAtrial fibrillation (AF) is the most frequent arrhythmic disease in humans, which leads to thrombus formation in the left atrial appendage and stroke through peripheral embolization. Depending on their origin, large extracellular vesicles (lEVs) can exert pro-coagulant functions. In the present study, we investigated how different types of AF influence the levels of large EV subtypes in three distinct atrial localizations. Blood samples were collected from the right and left atrium and the left atrial appendage of 58 patients. 49% of the patients had permanent AF, 34% had non-permanent AF, and 17% had no history of AF. Flow cytometric analysis of the origin of the lEVs showed that the proportion of platelet-derived lEVs in the left atrial appendage was significantly higher in permanent AF patients compared to non-permanent AF. When we grouped patients according to their current heart rhythm, we also detected significantly higher levels of platelet-derived lEVs in the left atrial appendage (LAA) in patients with atrial fibrillation. In vitro studies revealed, that platelet activation with lipopolysaccharide (LPS) leads to higher levels of miR-222-3p and miR-223-3p in platelet-derived lEVs. Treatment with lEVs from LPS- or thrombin-activated platelets reduces the migration of endothelial cells in vitro. These results suggest that permanent atrial fibrillation is associated with increased levels of platelet-derived lEVs in the LAA, which are potentially involved in LAA thrombus formation.

Sign in / Sign up

Export Citation Format

Share Document