AFibNet: An Implementation of Atrial Fibrillation Detection With Convolutional Neural Network

Abstract

Background: Generalization model capacity of deep learning (DL) approach for atrial fibrillation (AF) detection remains lacking. It can be seen from previous researches, the DL model formation used only a single frequency sampling of the specific device. Besides, each electrocardiogram (ECG) acquisition dataset produces a different length and sampling frequency to ensure sufficient precision of the R-R intervals to determine the Heart Rate Variability (HRV). An accurate HRV is the gold standard for predicting the AF condition. Hence, we propose a DL approach to analyze massive amounts of ECG raw data in a broad range of devices to overcome a current challenge.Results: This paper demonstrates powerful results for end-to-end implementation of AF detection based on a convolutional neural network (AFibNet). The method used a single learning system without considering the variety of signal lengths and frequency samplings. For implementation, the AFibNet is processed with a computational cloud-based DL approach. This study utilized a one-dimension convolutional neural networks (1D-CNNs) model for 11,842 subjects. It was trained and validated with 8,232 records based on three datasets and tested with 3,610 records based on eight datasets. The predicted results, when compared with the diagnosis results indicated by human practitioners, showed a 99.80% accuracy, sensitivity, and specificity. When tested with unseen data, the AF detection reaches 98.94% accuracy, 98.97% sensitivity, and 98.97% specificity in 0.02 seconds for one instance when processed in theDL-Cloud System.Conclusions: These findings demonstrate that the proposed model approach can used in a broad range of devices and validated to unknown data to derive feature maps and reliably detect the AF periods. We have found that our cloud-DL system is suitable for practical deployment.

Automated atrial fibrillation recognition in 12-lead electrocardiographic records: a signal to image and transfer learning approach: A case-control accuracy study

“…Previous studies have documented the identification of AF in ECG data using different approaches. For example, Tutuko et al [15] proposed a 1D-DCNN that is able to detect AF in unseen data, differentiating it from normal recordings, with an accuracy of 98.8%. A multilabel (AF vs. normal vs. other arrhythmia) 1D-DCNN reported an AF detection accuracy of 82% [16].…”

Artificial intelligence for the detection, prediction, and management of atrial fibrillation

“…Single-lead ECG may be integrated into a smartwatch wristband, making it widely available, and one study found a high sensitivity (97.7%) albeit a low PPV (40%) validated against an implantable cardiac monitor [26]. Cloud-based AI analysis of wearable ECG may allow continuous updates to the algorithm and facilitate contact with healthcare personnel [27]. Deep learning appears to be a stronger tool than classical machine learning for single-lead ECG analysis including detection of AF [28], and single-lead ECG may play a prominent role in AF detection in the future.…”

Electrocardiogram Analysis of Post-Stroke Elderly People Using One-dimensional Convolutional Neural Network Model with Gradient-weighted Class Activation Mapping

“…A DNN architecture is called convolutional neural network (CNN), which is particularly designed for processing time series data such as electrocardiograms (ECGs). Previous studies had used CNNs to detect ECG characteristic waveforms [9], arrhythmia [10,11], atrial fibrillation [12,13], hypertrophic cardiomyopathy [14], and myocardial infarction [15].…”