Cardiac Arrhythmia Classification Using Advanced Deep Learning Techniques on Digitized ECG Datasets.

ECG classification or heartbeat classification is an extremely valuable tool in cardiology. Deep learning-based techniques for the analysis of ECG signals assist human experts in the timely diagnosis of cardiac diseases and help save precious lives. This research aims at digitizing a dataset of images of ECG records into time series signals and then applying deep learning (DL) techniques on the digitized dataset. State-of-the-art DL techniques are proposed for the classification of the ECG signals into different cardiac classes. Multiple DL models, including a convolutional neural network (CNN), a long short-term memory (LSTM) network, and a self-supervised learning (SSL)-based model using autoencoders are explored and compared in this study. The models are trained on the dataset generated from ECG plots of patients from various healthcare institutes in Pakistan. First, the ECG images are digitized, segmenting the lead II heartbeats, and then the digitized signals are passed to the proposed deep learning models for classification. Among the different DL models used in this study, the proposed CNN model achieves the highest accuracy of ∼92%. The proposed model is highly accurate and provides fast inference for real-time and direct monitoring of ECG signals that are captured from the electrodes (sensors) placed on different parts of the body. Using the digitized form of ECG signals instead of images for the classification of cardiac arrhythmia allows cardiologists to utilize DL models directly on ECG signals from an ECG machine for the real-time and accurate monitoring of ECGs.

Mechanical dyssynchrony: How do we measure it, what it means, and what we can do about it.

Left ventricular mechanical dyssynchrony (LVMD) is defined by a difference in the timing of mechanical contraction or relaxation between different segments of the left ventricle (LV). Mechanical dyssynchrony is distinct from electrical dyssynchrony as measured by QRS duration and has been of increasing interest due to its association with worse prognosis and potential role in patient selection for cardiac resynchronization therapy (CRT). Although echocardiography is the most used modality to assess LVMD, some limitations apply to this modality. Compared to echo-based modalities, nuclear imaging by gated single-photon emission computed tomography (GSPECT) myocardial perfusion imaging (MPI) has clear advantages in evaluating systolic and diastolic LVMD. GSPECT MPI can determine systolic and diastolic mechanical dyssynchrony by the variability in the timing in which different LV segments contract or relax, which has prognostic impact in patients with coronary artery disease and heart failure. As such, by targeting mechanical dyssynchrony instead of electrical dyssynchrony, GSPECT MPI can potentially improve patient selection for CRT. So far, few studies have investigated the role of diastolic dyssynchrony, but recent evidence seems to suggest high prevalence and more prognostic impact than previously recognized. In the present review, we provide an oversight of mechanical dyssynchrony.

A Metaheuristic Optimization Approach for Parameter Estimation in Arrhythmia Classification from Unbalanced Data.

The electrocardiogram records the heart's electrical activity and generates a significant amount of data. The analysis of these data helps us to detect diseases and disorders via heart bio-signal abnormality classification. In unbalanced-data contexts, where the classes are not equally represented, the optimization and configuration of the classification models are highly complex, reflecting on the use of computational resources. Moreover, the performance of electrocardiogram classification depends on the approach and parameter estimation to generate the model with high accuracy, sensitivity, and precision. Previous works have proposed hybrid approaches and only a few implemented parameter optimization. Instead, they generally applied an empirical tuning of parameters at a data level or an algorithm level. Hence, a scheme, including metrics of sensitivity in a higher precision and accuracy scale, deserves special attention. In this article, a metaheuristic optimization approach for parameter estimations in arrhythmia classification from unbalanced data is presented. We selected an unbalanced subset of those databases to classify eight types of arrhythmia. It is important to highlight that we combined undersampling based on the clustering method (data level) and feature selection method (algorithmic level) to tackle the unbalanced class problem. To explore parameter estimation and improve the classification for our model, we compared two metaheuristic approaches based on differential evolution and particle swarm optimization. The final results showed an accuracy of 99.95%, a F1 score of 99.88%, a sensitivity of 99.87%, a precision of 99.89%, and a specificity of 99.99%, which are high, even in the presence of unbalanced data.

Arrhythmias in hematopoietic stem cell transplantation: A systematic review and meta-analysis.

BACKGROUND: There are controversial data regarding the relationship between hematopoietic stem cell transplantation and arrhythmias. This meta-analysis was performed to evaluate the incidence of arrhythmias in patients following hematopoietic stem cell transplantation (HSCT). METHODS: A literature search was conducted utilizing MEDLINE, EMBASE, and Cochrane Databases from inception through April 2019. Pooled incidence with 95% confidence interval (CI) were calculated using random-effects meta-analysis. The protocol for this meta-analysis is registered with PROSPERO (International Prospective Register of Systematic Reviews; no. CRD42019131833). RESULTS: Thirteen studies consisting of 10,587 patients undergoing HSCT were enrolled in this systematic review. Overall, the pooled estimated incidence of all types of arrhythmias following HSCT was 7.2% (95% CI: 4.9%-10.5%). With respect to the most common type of arrhythmia, the pooled estimated incidence of atrial fibrillation/atrial flutter (AF/AFL) within 30 days following HSCT was 4.2% (95% CI: 1.7%-9.6%). Egger's regression test demonstrated no significant publication bias in this meta-analysis of post-HSCT arrhythmia incidence. CONCLUSION: The overall estimated incidence of arrhythmias following HSCT was 7.2%. Future large scale studies are needed to further elucidate the significance and clinical impact of arrhythmias in post-HSCT patients.

A Decision Tree-Based Survival Analysis of Patients with a History of Inappropriate Implantable Cardioverter-Defibrillator Therapy.

Implantable cardioverter-defibrillators (ICDs) improve survival in patients who are at risk of sudden death. However, inappropriate therapy is commonly given to ICD recipients, and this situation may be associated with an increased risk of death. This study aimed to construct a risk stratification scheme by using decision tree analysis in patients who received inappropriate ICD therapy.Mortality was calculated from a retrospective data analysis of a multicenter cohort involving 417 ICD recipients. Inappropriate therapy was defined as therapy for nonventricular arrhythmias, including sinus tachycardia, supraventricular tachycardia, atrial fibrillation/flutter, oversensing, and lead failure. Inappropriate therapy included antitachycardia pacing, cardioversion, and defibrillation. The prognostic factors were identified by a Cox proportional hazards regression analysis, and we constructed a decision tree.During an average follow-up of 5.2 years, 48 patients (12%) had all-cause death. A multivariate Cox hazard model revealed that the age (hazard ratio [HR] 1.06, P < 0.001), ln B-type natriuretic peptide (BNP) (HR 1.47, P = 0.02), nonsinus rhythm at implantation (HR 2.70, P < 0.05), and inappropriate therapy occurring during sedentary/awake conditions (HR 3.51, P = 0.001) correlated with an increased risk of mortality. An inappropriate therapy due to abnormal sensing (HR 0.16, P = 0.04) decreased the risk of mortality. Furthermore, a decision tree analysis stratified the patients well by using 4 covariates: BNP, activity at the time of inappropriate therapy, mechanism of inappropriate therapy, and baseline rhythm at ICD implantation (log-rank test, P < 0.0001).We identified the predictors of mortality in inappropriate ICD therapy recipients and constructed a risk stratification scheme by using decision tree analysis.

An Improved Convolutional Neural Network Based Approach for Automated Heartbeat Classification.

With age, our blood vessels are prone to aging, which induces cardiovascular disease. As an important basis for diagnosing heart disease and evaluating heart function, the electrocardiogram (ECG) records cardiac physiological electrical activity. Abnormalities in cardiac physiological activity are directly reflected in the ECG. Thus, ECG research is conducive to heart disease diagnosis. Considering the complexity of arrhythmia detection, we present an improved convolutional neural network (CNN) model for accurate classification. Compared with the traditional machine learning methods, CNN requires no additional feature extraction steps due to the automatic feature processing layers. In this paper, an improved CNN is proposed to automatically classify the heartbeat of arrhythmia. Firstly, all the heartbeats are divided from the original signals. After segmentation, the ECG heartbeats can be inputted into the first convolutional layers. In the proposed structure, kernels with different sizes are used in each convolution layer, which takes full advantage of the features in different scales. Then a max-pooling layer followed. The outputs of the last pooling layer are merged and as the input to fully-connected layers. Our experiment is in accordance with the AAMI inter-patient standard, which included normal beats (N), supraventricular ectopic beats (S), ventricular ectopic beats (V), fusion beats (F), and unknown beats (Q). For verification, the MIT arrhythmia database is introduced to confirm the accuracy of the proposed method, then, comparative experiments are conducted. The experiment demonstrates that our proposed method has high performance for arrhythmia detection, the accuracy is 99.06%. When properly trained, the proposed improved CNN model can be employed as a tool to automatically detect different kinds of arrhythmia from ECG.

[Automatic classification method of arrhythmia based on discriminative deep belief networks].

Existing arrhythmia classification methods usually use manual selection of electrocardiogram (ECG) signal features, so that the feature selection is subjective, and the feature extraction is complex, leaving the classification accuracy usually affected. Based on this situation, a new method of arrhythmia automatic classification based on discriminative deep belief networks (DDBNs) is proposed. The morphological features of heart beat signals are automatically extracted from the constructed generative restricted Boltzmann machine (GRBM), then the discriminative restricted Boltzmann machine (DRBM) with feature learning and classification ability is introduced, and arrhythmia classification is performed according to the extracted morphological features and RR interval features. In order to further improve the classification performance of DDBNs, DDBNs are converted to deep neural network (DNN) using the Softmax regression layer for supervised classification in this paper, and the network is fine-tuned by backpropagation. Finally, the Massachusetts Institute of Technology and Beth Israel Hospital Arrhythmia Database (MIT-BIH AR) is used for experimental verification. For training sets and test sets with consistent data sources, the overall classification accuracy of the method is up to 99.84% ± 0.04%. For training sets and test sets with inconsistent data sources, a small number of training sets are extended by the active learning (AL) method, and the overall classification accuracy of the method is up to 99.31% ± 0.23%. The experimental results show the effectiveness of the method in arrhythmia automatic feature extraction and classification. It provides a new solution for the automatic extraction of ECG signal features and classification for deep learning.

Clinical impact, safety, and accuracy of the remotely monitored implantable loop recorder Medtronic Reveal LINQTM.

Aims: Implantable loop recorders (ILR) are indicated in a variety of clinical situations when extended cardiac rhythm monitoring is needed. We aimed to assess the clinical impact, safety, and accuracy of the new Medtronic Reveal LINQTM ILR that can be inserted outside the electrophysiology (EP) laboratory and remotely monitored. Methods and results: All 154 consecutive patients (100 males, 63 ± 15 year-old) who received the Reveal LINQTM ILR during the period July 2014-June 2016 were enrolled. The device was implanted in a procedure room and all patients where provided with the MyCareLinkTM remote monitoring system. Data were reviewed every working day via the Carelink® web system by a specialist nurse who, in case of significant events, consulted an electrophysiologist. During a mean follow-up of 12.1 (6.7-18.4) months (range 2-24 months), a diagnosis was made in 99 (64%) patients and in 60 (39%) ≥1 therapeutic interventions were established following recording of arrhythmias. In 26 of these 60 patients, remote monitoring prompted therapeutic interventions following asymptomatic arrhythmic events 3.8 months before the next theoretical scheduled in-office data download. False bradycardia detection for undersensing occurred in 44 (29%) patients and false tachycardia detection for oversensing in 4 (3%). One patient experienced skin erosion requiring explantation and none suffered from infection. Conclusion: The remote monitoring feature of the Reveal LINQTM allowed earlier diagnosis of asymptomatic but serious arrhythmias in a significant proportion of patients. Implantation of the device outside the EP laboratory appeared safe. However, R-wave undersensing and consequent false recognition of bradyarrhythmias remains a clinically important technical issue.

Magnetocardiographic classification and non-invasive electro-anatomical imaging of outflow tract ventricular arrhythmias in recreational sport activity practitioners.

Ventricular arrhythmias (VAs) with left bundle-branch-block and inferior axis morphology (LBBB-IA), suggestive of outflow tract (OT) origin, are a challenge in sports medicine because they can be benign or expression of a silent cardiomyopathy. Non-invasive classification is essential to plan ablation strategy if required. We aimed to evaluating magnetocardiographic (MCG) discrimination of OT-VAs site of origin (SoO). MCG and ECG data of 26 sports activity practitioners, with OT-VAs were analyzed. OT-VAs-SoO was classified with discriminant analysis (DA) of 8 MCG parameters and with invasively-validated ECG algorithms. MCG inverse source-localization merged with magnetic resonance (CMR) provided three-dimensional electro-anatomical imaging (MCG 3D-EAI). ECG classification was univocal in 73%. MCG-DA differentiated right ventricular OT from aortic sinus cusp VAs, with 94.7% accuracy. MCG 3D-EAI confirmed OT-VAs-SoO in CMR images. In cases undergoing ablation, MCG 3D-EAI was confirmed by CARTO 3D-EAI. MCG-DA improves non-invasive classification of OT-VAs-SoO. Further comparison with interventional results is required.

[Severe cardiac arrhythmias]. / Les troubles du rythme cardiaque sévères.

Cardiac arrhythmias are a common cause of admission in the emergency department. Among these, atrio-ventricular conductive disorders and malignant ventricular arrhythmias are among the most severe, requiring prompt and appropriate management to ensure the best prognosis. Knowledge of the pathophysiology and etiology causing these arrhythmias is mandatory in order to understand its management, acute and chronic, and to facilitate the dialogue between emergency physicians and cardiologists.

Les arythmies cardiaques sont une cause fréquente d'admission aux urgences. Parmi celles-ci, les troubles conductifs atrio-ventriculaires et les arythmies ventriculaires et supraventriculaires malignes sont à classer parmi les plus sévères. Elles nécessitent une prise en charge rapide et appropriée afin de garantir le meilleur pronostic possible aux patients. La connaissance de la physiopathologie et des étiologies engendrant ce type d'arythmie est nécessaire afin d'en comprendre la prise en charge, aiguë et chronique, et de faciliter le dialogue entre urgentistes et cardiologues.

Normality that is abnormal.

Supernormal conduction is defined as better-than-expected conduction in patients with depressed conduction during a short interval in the ventricular cycle. It is mainly observed in long-duration electrocardiogram (ECG) assessments. Its occurrence during 12-lead ECG is uncommon and its interpretation demands knowledge on electrophysiological alterations that are hard to understand. By reporting this case we aim to propose a rationale sequence that should be considered when facing an ECG with these same features, which would enable a greater accuracy to make a definitive diagnosis.

Arrhythmias in Adult Congenital Patients With Bodily Isomerism.

There are an increasing number of adults with congenital heart disease, some of whom have bodily isomerism. Bodily isomerism or heterotaxy is a unique clinical entity associated with congenital malformations of the heart which further increases the risk for future cardiovascular complications. We aimed to investigate the frequency of arrhythmias in adults with bodily isomerism. We utilized the 2012 iteration of the Nationwide Inpatient Sample to identify adult inpatient admissions associated with arrhythmias in patients with isomerism. Data regarding demographics, comorbidities, and various procedures were collected and compared between those with and without isomerism. A total of 6,907,109 admissions were analyzed with a total of 861 being associated isomerism. The frequency of arrhythmias was greater in those with isomerism (20.8 vs. 15.4 %). Those with isomerism were also more five times more likely to undergo invasive electrophysiology studies. Length and cost of hospitalization for patients with arrhythmias also tended to be greater in those with isomerism. Mortality did not differ between the two groups. Arrhythmias are more prevalent in those with isomerism, with a majority of arrhythmias in isomerism being atrial. Those with isomerism and arrhythmias also tended to have greater length and cost of hospitalization.

An APN model for Arrhythmic beat classification.

MOTIVATION: Changes in the normal rhythm of a human heart may result in different cardiac arrhythmias, which may be immediately fatal or cause irreparable damage to the heart sustained over long periods of time. Therefore, the ability to automatically identify arrhythmias from ECG recordings is important for clinical diagnosis and treatment. In this article, classification by using associative Petri net (APN) for personalized ECG-arrhythmia-pattern identification is proposed for the first time in literature. RESULTS: A rule-based classification model and reasoning algorithm of APN are created for ECG arrhythmias classification. The performance evaluation using MIT-BIH arrhythmia database shows that our approach compares well with other reported studies.

A comparison of genetic findings in sudden cardiac death victims and cardiac patients: the importance of phenotypic classification.

Sudden cardiac death (SCD) is responsible for a large proportion of non-traumatic, sudden and unexpected deaths in young individuals. Sudden cardiac death is a known manifestation of several inherited cardiac diseases. In post-mortem examinations, about two-thirds of the SCD cases show structural abnormalities at autopsy. The remaining cases stay unexplained after thorough investigations and are referred to as sudden unexplained deaths. A routine forensic investigation of the SCD victims in combination with genetic testing makes it possible to establish a likely diagnosis in some of the deaths previously characterized as unexplained. Additionally, a genetic diagnose in a SCD victim with a structural disease may not only add to the differential diagnosis, but also be of importance for pre-symptomatic family screening. In the case of SCD, the optimal establishment of the cause of death and management of the family call for standardized post-mortem procedures, genetic screening, and family screening. Studies of genetic testing in patients with primary arrhythmia disorders or cardiomyopathies and of victims of SCD presumed to be due to primary arrhythmia disorders or cardiomyopathies, were systematically identified and reviewed. The frequencies of disease-causing mutation were on average between 16 and 48% in the cardiac patient studies, compared with ∼10% in the post-mortem studies. The frequency of pathogenic mutations in heart genes in cardiac patients is up to four-fold higher than that in SCD victims in a forensic setting. Still, genetic investigation of SCD victims is important for the diagnosis and the possible investigation of relatives at risk.

Abnormal electrocardiographic QRS transition zone and risk of mortality in individuals free of cardiovascular disease.

AIMS: We examined the prognostic significance of abnormal electrocardiographic QRS transition zone (clockwise and counterclockwise horizontal rotations) in individuals free of cardiovascular disease (CVD). METHODS AND RESULTS: A total of 5541 adults (age 53 ± 10.4 years, 54% women, 24% non-Hispanic black, 25% Hispanic) without CVD or any major electrocardiogram (ECG) abnormalities from the US Third National Health and Nutrition Examination Survey were included in this analysis. Clockwise and counterclockwise horizontal rotations were defined from standard 12-lead ECG using Minnesota ECG Classification. Mortality and cause of death were assessed through 2006. At baseline, 282 participants had clockwise rotation and 3500 had counterclockwise rotation. During a median follow of 14.6 years, 1229 deaths occurred of which 415 were due to CVD. In multivariable-adjusted Cox proportional hazard analysis and compared with normal rotation, clockwise rotation was significantly associated with increased risk of all-cause mortality {hazard ratio (HR) [95% confidence interval (CI)]: 1.43 (1.15-1.78); P = 0.002} and CVD mortality [HR (95% CI): 1.61 (1.09, 2.37) P = 0.016]. In contrast, counterclockwise rotation was associated with significantly lower risk of all-cause mortality [HR (95% CI): 0.86 (0.76, 0.97); P = 0.017] and non-significant association with CVD mortality [HR (95% CI): 1.07 (0.86, 1.33); P = 0.549]. These results were consistent in subgroup analysis stratified by age, sex, and race. CONCLUSION: In a diverse community-based population free of CVD and compared with normal rotation, clockwise rotation was associated with increased risk of all-cause and CVD mortality while counterclockwise rotation was associated with lower risk of all-cause mortality and non-significant association with CVD mortality. These findings call for attention to these often neglected ECG markers, and probably call for revising the current definition of normal rotation.

Diagnostic accuracy of pace spikes in the electrocardiogram to diagnose paced rhythm.

OBJECTIVE: To determine how often cardiac resynchronization therapy (CRT) pacing systems generate visible pace spikes in the electrocardiogram (ECG). METHODS: In 46 patients treated with CRT pacing systems, we recorded ECGs during intrinsic rhythm, atrial pacing and ventricular pacing. ECGs were analysed for atrial and ventricular pace spikes by two experienced ECG readers blinded to the pacing therapy and to the study purpose. RESULTS: Atrial pacing generated visible pace spikes in less than 70% of the ECGs, whereas ventricular pacing generated visible pace spikes in about 90% of ECGs. The sensitivity of manual ECG interpretation for pace spikes was low for atrial pacing (Reader 1: 0.62 [95% confidence interval (CI) 0.50-0.74]; Reader 2: 0.65 [95% CI 0.53-0.77]) and moderate for ventricular pacing (Reader 1: 0.88 [95% CI 0.81-0.93]; Reader 2: 0.93 [95% CI 0.87-0.97]). CONCLUSIONS: In patients with CRT pacing systems, the absence of visible pace spikes in the ECG does not rule out paced rhythm.

Arrhythmias in the paediatric intensive care unit: a prospective study of the rates and predictors of arrhythmias in children without underlying cardiac disease.

OBJECTIVE: Arrhythmias are common in patients admitted to the paediatric intensive care unit. We sought to identify the rates of occurrence and types of arrhythmias, and determine whether an arrhythmia was associated with illness severity and paediatric intensive care unit length of stay. DESIGN: This is a prospective, observational study of all patients admitted to the paediatric intensive care unit at the Children's Hospital at Montefiore from March to June 2012. Patients with cardiac disease or admitted for the treatment of primary arrhythmias were excluded. Clinical and laboratory data were collected and telemetry was reviewed daily. Tachyarrhythmias were identified as supraventricular tachycardia, ventricular tachycardia, and arrhythmias causing haemodynamic compromise or for which an intervention was performed. RESULTS: A total of 278 patients met the inclusion criteria and were analysed. There were 97 incidences of arrhythmia in 53 patients (19%) and six tachyarrhythmias (2%). The most common types of arrhythmias were junctional rhythm (38%), premature atrial contractions (24%), and premature ventricular contractions (22%). Tachyarrhythmias included three supraventricular tachycardia (50%) and three ventricular tachycardia (50%). Of the six tachyarrhythmias, four were related to placement or migration of central venous lines and two occurred during aminophylline infusion. Patients with an arrhythmia had longer duration of mechanical ventilation and paediatric intensive care unit stay (p<0.001). In multivariate analysis, central venous lines (odds ratio 3.1; 95% confidence interval 1.3-7.2, p=0.009) and aminophylline use (odds ratio 5.1; 95% confidence interval 1.7-14.9, p=0.003) were independent predictors for arrhythmias. CONCLUSIONS: Arrhythmias were common in paediatric intensive care unit patients (19%), although tachyarrhythmias occurred rarely (2%). Central venous lines and use of aminophylline were identified as two clinical factors that may be associated with development of an arrhythmia.

[Congenital Cardiac Conduction Abnormalities].

Abnormalities in cardiac conduction can occur due to a variety of factors. So called "idiopathic", conduction system degeneration develops without evident causes and may have hereditary basis. In the majority of cases it has no clinical manifestation, do not require treatment and have overall good prognosis. In this review we focus on congenital complete atrioventricular block and progressive cardiac conduction defect - rare but malignant and potentially lethal conditions that can be caused by genetic mutations and may be isolated or associated with structural heart disease. Cardiac involvement is relatively common in rare hereditary diseases - myodystrophies and mitochondrial cytopathies. Conduction abnormalities are among the most severe manifestations that may determine prognosis in these rare genetic disorders. These conditions deserve special consideration because of rapid progression of conduction defects and high prevalence of sudden cardiac death if no appropriate treatment applied.

[CHRONIC DESTRUCTIVE PULMONARY TUBERCULOSIS WITH DISTURBANCE OF RHYTHM AND CONDUCTION OF THE HEART].

60 patients were observed with chronic destructive pulmonary tuberculosis which was diagnosed rhythm disturbance and conduction of the heart. It should be noted that the most frequently occurred among patients with heart rhythm disturbances in the form of sinus tachycardia (27.4%) and sinus arrhythmia (24.2%). Almost half of the patients (48, 1%) in the electrocardiogram (EKQ) were diagnosed disturbance of intraventricular conduction. The disturbance of atrioventricular conduction noted in 33.3% of cases. As a result of Halter ECG monitoring sunpeventricular disturbances are diagnosed most frequently and take place in 63,6% of cases. Ventricular arrhythmias were diagnosed in 12% of patients. Conduction of the first of degree atrioventricular block transitory character was noted in 55.6% of cases. Intraventricular blockades of Hiss right branch were determined in 44.4% of cases. The most frequently (41.6%) among the associated with rhythm disturbances and cardiac conduction diagnosed a rare combination of supraventricular monotypic extrasystole with first class ventricular extrasystole.