Search

With the rapid development of Internet of Things (IoT), it becomes convenient to use mobile devices to remotely monitor the Physiological signals (e.g., Arrhythmia diseases) of patients with chronic diseases (e.g., cardiovascular diseases (CVDs)). High classification accuracy of inter-patient ECGs is extremely important for diagnosing Arrhythmia. The Supraventricular ectopic beat (S) is especially difficult to be classified. It is often misclassified as Normal (N) or Ventricular ectopic beat (V). Class imbalance is another common and important problem in electronic health (eHealth), as abnormal samples (i.e., samples of specific diseases) are usually far less than normal samples. To solve these problems, we propose a sustainable deep learning-based heartbeat classification system, called BeatClass. It contains three main components: two stacked bidirectional Long Short-Term Memory Networks (Bi-LSTMs), called Rist and Morst, and a Generative Adversarial Network (GAN), called MorphGAN. Rist first classifies the heartbeats into five common Arrhythmia classes. The heartbeats classified as S and V by Rist are further classified by Morst to improve the classification accuracy. MorphGAN is used to augment the morphological and contextual knowledge of heartbeats in infrequent classes. In the experiment, BeatClass is compared with several state-of-the-art works for inter-patient arrhythmia classification. The F1-scores of classifying N, S and V heartbeats are 0.6%, 16.0% and 1.8% higher than the best baseline method. The experiment result demonstrates that taking multiple classification models to improve classification results step-by-step may significantly improve the classification performance. We also evaluate the classification sustainability of BeatClass. Based on different physical signal datasets, a trained BeatClass can be updated to classify heartbeats with different sampling rates. At last, an engineering application indicates that BeatClass can promote the sustainable development of IoT-based eHealth.

Background and Objective: Ectopic beats originating from the pulmonary vein (PV) trigger atrial fibrillation (AF). The purpose of this study was to clarify the electrophysiological determinant of AF initiation from the PVs. Methods: Pacing studies were performed with a single extra stimulus mimicking an ectopic beat in the left superior pulmonary veins (LSPVs) in 62 patients undergoing AF ablation. Inducibility of AF, effective refractory period (ERP) and conduction properties within the PVs were analyzed. Results: A single extra stimulus in LSPV induced AF in 20 patients (32% of all patients) at the mean coupling interval (CI) of 172 ms. A CI-dependent anisotropic conduction at the AF onset was visualized in a 3D-mapping. Onset of AF was site-specific with reproducibility in each individual. Mean ERP in LSPV in the AF inducible group was shorter than that in the AF non-inducible group (182 ± 55 ms vs 254 ± 51 ms, P

This paper presents a system for electrocardiogram (ECG) delineation and arrhythmia classification. The proposed system consists of a front-end integrated circuit, a delineation algorithm implemented on an FPGA board, and an arrhythmia classification algorithm. The front-end circuit applies a ternary second-order Delta modulator to measure the slope variation of the input analog ECG signal. The circuit converts the analog inputs into a pulse density modulated bitstream, whose pulse density is proportional to the slope variation of the input analog signal regardless of the instantaneous amplitude. The front-end chip can detect the minimum slope variation of 3.2 mV/ms 2 within a 3 ms timing error. The front-end integrated circuit was fabricated with a 180 nm CMOS process occupying a 0.25mm 2 area with a 151 nW power consumption at the sampling rate of 1 kS/s. Based on the slope variation obtained from the front-end circuit, a delineation algorithm is designed to detect fiducial points in the ECG waveform. The delineation algorithm was tested on a Spartan-6 FPGA. The delineation system can detect the intervals, slopes, and morphology of the QRS/PT waves and form a feature set that contains 22 features. Based on these features, a rotate linear kernel support vector machine (SVM) is applied for patient-specific arrhythmia classification of the ventricular ectopic beat (VEB), supraventricular ectopic beat (SVEB), and normal heartbeat. The performance of the proposed system is comparable to the recently published methods while providing a promising solution for the low-complexity implementation of future wearable ECG monitoring systems.

The regular monitoring and accurate diagnosis of arrhythmia are critically important, leading to a reduction in mortality rate due to cardiovascular diseases (CVD) such as heart stroke or cardiac arrest. This paper proposes a novel convolutional neural network (CNN) model for arrhythmia classification. The proposed model offers the following improvements compared with traditional CNN models. Firstly, the multi-channel model can concatenate spectral and spatial feature maps. Secondly, the structural unit is composed of a depthwise separable convolution layer followed by activation and batch normalization layers. The structural unit offers effective utilization of network parameters. Also, the optimization of hyperparameters is done using Hyperopt library, based on Sequential Model-Based Global Optimization algorithm (SMBO). These improvements make the network more efficient and accurate for arrhythmia classification. The proposed model is evaluated using tenfold cross-validation following both subject-oriented inter-patient and class-oriented intra-patient evaluation protocols. Our model achieved 99.48% and 99.46% accuracy in VEB (ventricular ectopic beat) and SVEB (supraventricular ectopic beat) class classification, respectively. The model is compared with state-of-the-art models and has shown significant performance improvement.

This paper develops an end-to-end ECG signal classification algorithm based on a novel segmentation strategy and 1D Convolutional Neural Networks (CNN) to aid the classification of ECG signals and alleviate the workload of physicians. The ECG segmentation strategy named R-R-R strategy (i.e., retaining ECG data between the R peaks just before and after the current R peak) is used for segmenting the original ECG data into segments to train and test the 1D CNN models. The novel strategy mimics physicians in scanning ECG to a greater extent, and maximizes the inherent information of ECG segments for diagnosis. The performance of the proposed end to end ECG signal classification algorithm was verified with the ECG signals from 48 records in the MIT-BIH arrhythmia database. When the heartbeat types were divided into the five classes recommended by clinicians, i.e., normal beat, left bundle branch block beat, right bundle branch block beat, premature ventricular contraction, and paced beat, the classification accuracy, the area under the curve (AUC), the sensitivity, and the F1-score achieved by the proposed model were 0.9924, 0.9994, 0.99 and 0.99, respectively. When the heartbeat types were divided into six classes recommended by clinicians, i.e., normal beat, left bundle branch block beat, right bundle branch block beat, premature ventricular contraction, paced beat and other beats, the beat classification accuracy, the AUC, the sensitivity, and the F1-score achieved by the model reached 0.9702, 0.9966, 0.97, and 0.97, respectively. When the heartbeat types were divided into five classes recommended by the Association for Advancement of Medical Instrumentation (AAMI), i.e., normal beat, supraventricular ectopic beat, ventricular ectopic beat, fusion beat, and unknown beat, the beat classification accuracy, the sensitivity, and the F1-score were 0.9745, 0.97, and 0.97, respectively. Experimental results show that the proposed method achieves better performance than the state-of-the-art methods.

Background: Premature ventricular contractions have been described in 1% of clinically normal people as detected by standard ECG and 40–75% of apparently healthy persons as detected by 24–48 hour ambulatory ECG recording (Holter). In the absence of structural heart diseases, it is thought to be relatively benign. However, PVC-induced cardiomyopathy has been a subject of great interest over the last decade. Objective: To evaluate subtle and early premature ventricular complex (PVC)–induced ventricular impairment in patients with frequent PVCs and without structural heart disease by real-time 3-dimensional speckle tracking echocardiography (RT3D-STE). Patients and Methods: Forty patients were included in this study with asymptomatic frequent PVCs. Patients without structural heart disease was excluded. Detailed history, physical examination, full laboratory investigations, resting ECG, 24 hours Holter monitoring, conventional 2D echocardiography assessment of left ventricular ejection fraction (LVEF), strain imaging with measurement of LV-GLS, 2D speckle tracking and RT3D-STE (global longitudinal, area, radial, and circumferential strain (GLS, GAS, GRS, and GCS)). The study was performed at Islamic Cardiac Centre of Cardiology, Al-Azhar University during the period from September 2019 to April 2020. Results: RT3D-STE modalities detectap early subtle LV dysfunction better than conventional echocardiographic and showed a significant correlation with PVCs burden, origin, focality, interpolated PVCs and duration of the complex of ectopic beat. Conclusion: Frequent premature ventricular complexes could produce LV dysfunction and cardiomyopathy independently of any pre-existing underlying structure heart disease and early detection of subtle LV dysfunction can be done by RT3D-STE modalities even in asymptomatic patient.

Real-time wearable electrocardiogram monitoring sensor is one of the best candidates in assisting cardiovascular disease diagnosis. In this paper, we present a novel real-time machine learning system for Arrhythmia classification. The system is based on the parallel Delta modulation and QRS/PT wave detection algorithms. We propose a patient dependent rotated linear-kernel support vector machine classifier that combines the global and local classifiers, with three types of feature vectors extracted directly from the Delta modulated bit-streams. The performance of the proposed system is evaluated using the MIT-BIH Arrhythmia database. According to the AAMI standard, two binary classifications are performed and evaluated, which are supraventricular ectopic beat (SVEB) versus the rest four classes, and ventricular ectopic beat (VEB) versus the rest. For SVEB classification, the preferred SkP-32 method's F1 score, sensitivity, specificity, and positive predictivity value are 0.83, 79.3%, 99.6%, and 88.2%, respectively, and for VEB classification, the numbers are 0.92%, 92.8%, 99.4%, and 91.6%, respectively. The results show that the performance of our proposed approach is comparable to that of published research. The proposed low-complexity algorithm has the potential to be implemented as an on-sensor machine learning solution.

Background Electrical storm (ES) is a life-threatening emergency in patients at high risk of ventricular tachycardia/ventricular fibrillation (VF), but the pathophysiology and molecular basis are poorly understood. Objective The purpose of this study was to explore the electrophysiological substrate for experimental ES. Methods A model was created by inducing chronic complete atrioventricular block in defibrillator-implanted rabbits, which recapitulates QT prolongation, torsades des pointes (TdP), and VF episodes. Results Optical mapping revealed island-like regions with action potential duration (APD) prolongation in the left ventricle, leading to increased spatial APD dispersion, in rabbits with ES (defined as ≥3 VF episodes/24 h). The maximum APD and its dispersion correlated with the total number of VF episodes in vivo. TdP was initiated by an ectopic beat that failed to enter the island and formed a reentrant wave and perpetuated by rotors whose centers swirled in the periphery of the island. Epinephrine exacerbated the island by prolonging APD and enhancing APD dispersion, which was less evident after late Na+ current blockade with 10 μM ranolazine. Nonsustained ventricular tachycardia in a non-ES rabbit heart with homogeneous APD prolongation resulted from multiple foci with an electrocardiographic morphology different from TdP driven by drifting rotors in ES rabbit hearts. The neuronal Na+-channel subunit NaV1.8 was upregulated in ES rabbit left ventricular tissues and expressed within the myocardium corresponding to the island location in optically mapped ES rabbit hearts. The NaV1.8 blocker A-803467 (10 mg/kg, intravenously) attenuated QT prolongation and suppressed epinephrine-evoked TdP. Conclusion A tissue island with enhanced refractoriness contributes to the generation of drifting rotors that underlies ES in this model. NaV1.8-mediated late Na+ current merits further investigation as a contributor to the substrate for ES.

Ectopic beats (EBs) are cellular arrhythmias that can trigger lethal arrhythmias. Simulations using biophysically-detailed cardiac myocyte models can reveal how model parameters influence the probability of these cellular arrhythmias, however such analyses can pose a huge computational burden. Here, we develop a simplified approach in which logistic regression models (LRMs) are used to define a mapping between the parameters of complex cell models and the probability of EBs (P(EB)). As an example, in this study, we build an LRM for P(EB) as a function of the initial value of diastolic cytosolic Ca2+ concentration ([Ca2+]iini), the initial state of sarcoplasmic reticulum (SR) Ca2+ load ([Ca2+]SRini), and kinetic parameters of the inward rectifier K+ current (IK1) and ryanodine receptor (RyR). This approach, which we refer to as arrhythmia sensitivity analysis, allows for evaluation of the relationship between these arrhythmic event probabilities and their associated parameters. This LRM is also used to demonstrate how uncertainties in experimentally measured values determine the uncertainty in P(EB). In a study of the role of [Ca2+]SRini uncertainty, we show a special property of the uncertainty in P(EB), where with increasing [Ca2+]SRini uncertainty, P(EB) uncertainty first increases and then decreases. Lastly, we demonstrate that IK1 suppression, at the level that occurs in heart failure myocytes, increases P(EB)., Author summary An ectopic beat is an abnormal cellular electrical event which can trigger dangerous arrhythmias in the heart. Complex biophysical models of the cardiac myocyte can be used to reveal how cell properties affect the probability of ectopic beats. However, such analyses can pose a huge computational burden. We develop a simplified approach that enables a highly complex biophysical model to be reduced to a rather simple statistical model from which the functional relationship between myocyte model parameters and the probability of an ectopic beat is determined. We refer to this approach as arrhythmia sensitivity analysis. Given the efficiency of our approach, we also use it to demonstrate how uncertainties in experimentally measured myocyte model parameters determine the uncertainty in ectopic beat probability. We find that, with increasing model parameter uncertainty, the uncertainty in probability of ectopic beat first increases and then decreases. In general, our approach can efficiently analyze the relationship between cardiac myocyte parameters and the probability of ectopic beats and can be used to study how uncertainty of these cardiac myocyte parameters influences the ectopic beat probability.