Smartwatch Electrocardiograms for Automated and Manual Diagnosis of Atrial Fibrillation: A Comparative Analysis of Three Models

Accuracy and interpretability of smartwatch electrocardiogram for early detection of atrial fibrillation: A systematic review and meta-analysis

Background

The prevalence of atrial fibrillation (AFib) continues to increase globally, posing a significant risk for serious cardiovascular complications, such as ischemic stroke and thromboembolism. Smartwatch single-lead electrocardiogram (ECG) can be a practical and accurate early detection tool for AFib.

Objective

The aim of this study was to fill the research gap in evaluating the accuracy and interpretability of smartwatch ECG for early AFib detection.

Methods

Data derived from indexed literature in the Scopus, Scilit, PubMed, Google Scholar, Web of Science, IEEE, and Cochrane Library databases (as of June 1, 2024) were systematically screened and extracted. The quantitative synthesis was performed using a two-level mixed-effects logistic regression model, as well as a proportional analysis with Freeman-Tukey double transformation on a restricted maximum-likelihood model.

Results

The sensitivity and specificity of smartwatch ECG in algorithmic readings were 86% and 94%, respectively. In manual readings, the sensitivity and specificity reached 96% and 95%, respectively. In a brand-specific subgroup analysis, the algorithmic reading reached a summary area under the curve (sAUC) of 96%, while another brand achieved the highest sAUC of 98% in manual reading. The level of manual interpretability was relatively high with Cohen's Kappa of 0.83, but 3% of ECG results were difficult to read manually.

Conclusion

This study shows that smartwatch ECG is able to detect AFib with high accuracy, especially through manual reading by trained medical personnel.

PROSPERO Registration

CRD42024548537 (May 29, 2024).

Home detection of atrial fibrillation using cardiac activity analysis: technologies available to the patient

INTRODUCTION

Atrial fibrillation (AF) is the most common cardiac arrhythmia, whose incidence and prevalence have increased over the last 20 years and will continue to increase over the next 30 years. It is characterized by irregular atrial activation, leading to complications as stroke and heart failure. Due to its intermittent and asymptomatic nature, diagnosing and monitoring AF is challenging but crucial for effective treatment and prevention of serious complications.

AREAS COVERED

This study reviews noninvasive medical devices available for home detection of AF by analyzing cardiac activity through ECG or photoplethysmography (PPG). The review covers the technologies underlying single-lead ECG acquisition and PPG sensors, and describes how these are used, also in combination, in home-use medical devices (including smartwatches and wristbands).

EXPERT OPINION

Single-lead ECG and PPG technologies in consumer electronics have revolutionized AF detection, making it more accessible and convenient for patients. Despite some limitations in signal quality and diagnostic scope, these devices offer significant benefits for early AF detection and management. The use of wearable devices, including smartwatches and wristbands, for heart activity monitoring represents a promising advancement in patient-lead healthcare, potentially leading to better outcomes through timely medical intervention and improved patient engagement in managing their condition.

Are Wearable ECG Devices Ready for Hospital at Home Application?

The increasing focus on improving care for high-cost patients has highlighted the potential of Hospital at Home (HaH) and remote patient monitoring (RPM) programs to optimize patient outcomes while reducing healthcare costs. This paper examines the role of wearable devices with electrocardiogram (ECG) capabilities for continuous cardiac monitoring, a crucial aspect for the timely detection and management of various cardiac conditions. The functionality of current wearable technology is scrutinized to determine its effectiveness in meeting clinical needs, employing a proposed ABCD guide (accuracy, benefit, compatibility, and data governance) for evaluation. While smartwatches show promise in detecting arrhythmias like atrial fibrillation, their broader diagnostic capabilities, including the potential for monitoring corrected QT (QTc) intervals during pharmacological interventions and approximating multi-lead ECG information for improved myocardial infarction detection, are also explored. Recent advancements in machine learning and deep learning for cardiac health monitoring are highlighted, alongside persistent challenges, particularly concerning signal quality and the need for further validation for widespread adoption in older adults and Hospital at Home settings. Ongoing improvements are necessary to overcome current limitations and fully realize the potential of wearable ECG technology in providing optimal care for high-risk patients.

Wearable Devices for Arrhythmia Detection: Advancements and Clinical Implications

Cardiac arrhythmias are a growing global health concern, and the need for accessible, continuous monitoring has driven rapid advancements in wearable technologies. This review explores the evolution, capabilities, and clinical impact of modern wearables for arrhythmia detection, including smartwatches, smart rings, ECG patches, and smart textiles. In light of the recent surge in commercially available wearables across all categories, this review offers a detailed comparative analysis of leading devices, evaluating cost, regulatory approval, model specifications, and system compatibility. Smartwatches and patches, in particular, show a strong performance in atrial fibrillation detection, with patches outperforming Holter monitors in long-term monitoring and diagnostic yield. This review highlights a paradigm shift toward patient-initiated diagnostics but also discusses challenges such as false positives, regulatory gaps, and healthcare integration. Overall, wearable devices hold significant promise for reshaping arrhythmia management through early detection and remote monitoring.

Comparison of diagnostic accuracy of electrocardiogram-based versus photoplethysmography-based smartwatches for atrial fibrillation detection: A Systematic Review and Meta-Analysis

Background

Atrial fibrillation (AF), the most prevalent cardiac arrhythmia, significantly affects morbidity and mortality, making early detection crucial for preventing stroke and heart failure. Recent advancements in wearable technology have introduced smartwatches as potential tools for continuous non-invasive AF detection.

Objective

This systematic review and meta-analysis aimed to evaluate and compare the diagnostic accuracy of electrocardiography (ECG)-and photoplethysmography (PPG)-based smartwatches in detecting AF.

Methodology

A comprehensive search was conducted on PubMed, Google Scholar, and other databases from 18 August to 23 September 2024, to fetch original studies that evaluated performance metrics of ECG and PPG smartwatches in AF detection. The obtained literature was screened according to preset inclusion and exclusion criteria. For included studies, the random-effects model was used to calculate their pooled sensitivity and specificity in AF detection using Jamovi 2.3.28 software. A significance threshold of P <0.05 was applied to all statistical analyses.

Results

Out of the 2564 studies screened, 25 met the inclusion criteria: 11 on PPG and 14 on ECG smartwatches. PPG smartwatches exhibited higher diagnostic performance with a pooled sensitivity of 97.4% (95% CI: 96.5-98.3) and specificity of 96.6% (95% CI: 94.9-98.3). Conversely, ECG smartwatches showed a pooled sensitivity of 83% (95% CI: 78-88) and specificity of 88.4% (95% CI: 84.5-92.2), lower than PPG smartwatches.

Conclusion

PPG-based smartwatches outperformed ECG-based devices in AF detection, offering higher sensitivity and specificity. Even though both modalities are effective in AF detection, the considerable variability in ECG smartwatch performance highlights the need for further research and standardization.

Design and validation of Withings ECG Software 2, a tiny neural network based algorithm for detection of atrial fibrillation

BACKGROUND

Atrial Fibrillation (AF) is the most common form of arrhythmia in the world with a prevalence of 1%-2%. AF is also associated with an increased risk of cardiovascular diseases (CVD), such as stroke, heart failure, and coronary artery diseases, making it a leading cause of death. Asymptomatic patients are a common case (30%-40%). This highlights the importance of early diagnosis or screening. Wearable and home devices offer new opportunities in this regard.

METHODS

We present WECG-SW2, a lightweight algorithm that classifies 30-second lead I ECG strips as 'NSR', 'AF', 'Other' or 'Noise'. By detecting the location of QRS complexes in the signal, the information can be organized into a low dimensionality input which is fed to a tiny Convolutional Neural Network (CNN) with only 3,633 parameters. This approach allows for the algorithm to run directly on the ECG acquisition devices, and improves accuracy by making the most out of the training set.

RESULTS

WECG-SW2 was evaluated on a database which combines three clinical studies sponsored by Withings with three hardware devices, as well as the MIT-BIH Arrhythmia Database. On the proprietary clinical database, the sensitivity and specificity of AF detection were 99.63% (95% CI: 99.15-99.84) and 99.85% (95% CI: 99.61-99.94), respectively, based on 4646 strips taken from 1441 participants. On the MIT-BIH Arrhythmia Database, the sensitivity and specificity were 99.87% (95% CI: 99.53, 99.98) and 100% (95% CI: 98.31, 100.0), respectively, across 2624 analyzed segments.

CONCLUSION

WECG-SW2 demonstrates high sensitivity and specificity in the detection of AF using a wide variety of ECG recording hardware. The binary of WECG-SW2 is available upon request to the corresponding author for research purposes.

Diagnostic Accuracy of Apple Watch Electrocardiogram for Atrial Fibrillation: A Systematic Review and Meta-Analysis

Background

Electrocardiography (ECG) is the gold standard for the diagnosis of atrial fibrillation (AF). Recently, smartwatches like the Apple Watch have emerged as a promising, user-friendly device for rapid detection and diagnosis of AF, but the reliability and diagnostic accuracy still remain controversial.

Objectives

The purpose of this study was to perform a systematic review and diagnostic test accuracy meta-analysis evaluating the diagnostic performance of the Apple Watch ECG in detecting AF.

Methods

The literature search was conducted on PubMed, Embase, and Cochrane Library through April 2024 for studies comparing the diagnostic accuracy of Apple Watch to standard 12-lead ECG. Statistical analysis was performed using R Software version 4.4.0 and OpenMeta[Analyst]. Pooled analyses of sensitivity, specificity, and area under the receiver operating characteristic curve were determined along with their 95% CIs. The quality of studies was analyzed using the QUADAS-2 tool.

Results

The meta-analysis included 11 studies comprising 4,241 participants. Their mean age was 62.56 ± 3.92 years, and 28% of the patients were females. The pooled sensitivity and specificity of the Apple Watch for detecting AF were 94.8% (95% CI: 91.7% to 96.8%; I2 = 67%) and 95% (95% CI: 88.6% to 97.8%; I2 = 88%), respectively. The area under the receiver operating characteristic curve was 0.96 (95% CI: 0.92-0.97).

Conclusions

The Apple Watch ECG carries high accuracy in detecting atrial fibrillation, providing a convenient diagnostic option for patients.

A novel recommender framework with chatbot to stratify heart attack risk

Cardiovascular diseases are a major cause of mortality and morbidity. Fast detection of life-threatening emergency events and an earlier start of the therapy would save many lives and reduce successive disabilities. Understanding the specific risk factors associated with heart attack and the degree of association is crucial in the clinical diagnosis. Considering the potential benefits of intelligent models in healthcare, many researchers have developed a variety of machine learning (ML)-based models to identify patients at risk of a heart attack. However, the common problem of previous works that used ML concepts was the lack of transparency in black-box models, which makes it difficult to understand how the model made the prediction. In this study, an automated smart recommender system (Explainable Artificial Intelligence) for heart attack prediction and risk stratification was developed. For the purpose, the CatBoost classifier was applied as the initial step. Then, the SHAP (SHapley Additive exPlanation) explainable algorithm was employed to determine reasons behind high or low risk classification. The recommender system can provide insights into the reasoning behind the predictions, including group-based and patient-specific explanations. In the final step, we integrated a Large Language Model (LLM) called BioMistral for chatting functionally to talk to users based on the model output as a digital doctor for consultation. Our smart recommender system achieved high accuracy in predicting a patient risk level with an average AUC of 0.88 and can explain the results transparently. Moreover, a Django-based online application that uses patient data to update medical information about an individual's heart attack risk was created. The LLM chatbot component would answer user questions about heart attacks and serve as a virtual companion on the route to heart health, our system also can locate nearby hospitals by applying Google Maps API and alert the users. The recommender system could improve patient management and lower heart attack risk while timely therapy aids in avoiding subsequent disabilities.

Rhythm-Ready: Harnessing Smart Devices to Detect and Manage Arrhythmias

PURPOSE OF REVIEW

To survey recent progress in the application of implantable and wearable sensors to detection and management of cardiac arrhythmias.

RECENT FINDINGS

Sensor-enabled strategies are critical for the detection, prediction and management of arrhythmias. In the last several years, great innovation has occurred in the types of devices (implanted and wearable) that are available and the data they collect. The integration of artificial intelligence solutions into sensor-enabled strategies has set the stage for a new generation of smart devices that augment the human clinician. Smart devices enhanced by new sensor technologies and Artificial Intelligence (AI) algorithms promise to reshape the care of cardiac arrhythmias.

Expert opinion on design and endpoints for studies on catheter ablation of atrial fibrillation

INTRODUCTION

Catheter ablation of atrial fibrillation (AF) is frequently studied in randomized trials, observational and registry studies. The aim of this expert opinion is to provide guidance for clinicians and industry regarding the development of future clinical studies on catheter ablation of AF, implement lessons learned from previous studies, and promote a higher degree of consistency across studies.

BACKGROUND

Studies on catheter ablation of AF may benefit from well-described definitions of endpoints and consistent methodology and documentation of outcomes related to efficacy, safety and cost-effectiveness. The availably of new, innovative technologies warrants further consideration about their application and impact on study design and the choice of endpoints. Moreover, recent insights gained from AF ablation studies suggest a reconsideration of some methodological aspects.

METHODS

A panel of clinical experts on catheter ablation of AF and designing and conducting clinical studies developed an expert opinion on the design and endpoints for studies on catheter ablation of AF. Discussions within the expert panel with the aim to reach consensus on predefined topics were based on outcomes reported in the literature and experiences from recent clinical trials.

RESULTS

A comprehensive set of recommendations is presented. Key elements include the documentation of clinical AF, medication during the study, repeated ablations and their effect on endpoint assessments, postablation blanking and the choice of rhythm-related and other endpoints.

CONCLUSION

This expert opinion provides guidance and promotes consistency regarding design of AF catheter ablation studies and identified aspects requiring further research to optimize study design and methodology.

CONDENSED ABSTRACT

Recent insights from studies on catheter ablation of atrial fibrillation (AF) and the availability of new innovative technologies warrant reconsideration of methodological aspects related to study design and the choice and assessment of endpoints. This expert opinion, developed by clinical experts on catheter ablation of AF provides a comprehensive set of recommendations related to these methodological aspects. The aim of this expert opinion is to provide guidance for clinicians and industry regarding the development of clinical studies, implement lessons learned from previous studies, and promote a higher degree of consistency across studies.

Artificial intelligence-enhanced patient evaluation: bridging art and science

The advent of digital health and artificial intelligence (AI) has promised to revolutionize clinical care, but real-world patient evaluation has yet to witness transformative changes. As history taking and physical examination continue to rely on long-established practices, a growing pipeline of AI-enhanced digital tools may soon augment the traditional clinical encounter into a data-driven process. This article presents an evidence-backed vision of how promising AI applications may enhance traditional practices, streamlining tedious tasks while elevating diverse data sources, including AI-enabled stethoscopes, cameras, and wearable sensors, to platforms for personalized medicine and efficient care delivery. Through the lens of traditional patient evaluation, we illustrate how digital technologies may soon be interwoven into routine clinical workflows, introducing a novel paradigm of longitudinal monitoring. Finally, we provide a skeptic's view on the practical, ethical, and regulatory challenges that limit the uptake of such technologies.

Artificial intelligence-based electrocardiogram analysis improves atrial arrhythmia detection from a smartwatch electrocardiogram

Aims

Smartwatch electrocardiograms (SW ECGs) have been identified as a non-invasive solution to assess abnormal heart rhythm, especially atrial arrhythmias (AAs) that are related to stroke risk. However, the performance of these tools is limited and could be improved with the use of deep neural network (DNN) algorithms, particularly for specific populations encountered in clinical cardiology practice.

Methods and results

A total of 400 patients from the electrophysiology department of one tertiary care hospital were included in two similar clinical trials (respectively, 200 patients per study). Simultaneous ECGs were recorded with the watch and a 12-lead recording system during consultation or before and after an electrophysiology procedure if any. The SW ECGs were processed by using the DNN and with the Apple watch ECG software (Apple app). Corresponding 12-lead ECGs (12L ECGs) were adjudicated by an expert electrophysiologist. The performance of the DNN was assessed vs. the expert interpretation of the 12L ECG, and inconclusive rates were reported. Overall, the DNN and the Apple app presented, respectively, a sensitivity of 91% [95% confidence interval (CI) 85-95%] and 61% (95% CI 44-75%) with a specificity of 95% (95% CI 91-97%) and 97% (95% CI 93-99%) when compared with the physician 12L ECG interpretation. The DNN was able to provide a diagnosis on 99% of ECGs, while the Apple app was able to classify only 78% of strips (22% of inconclusive diagnosis).

Conclusion

In this study, by including patients from a cardiology department, a DNN-based algorithm applied to an SW ECG provided an accurate diagnosis for AA detection on virtually all tracings, outperforming the SW algorithm.

Feasibility of remote monitoring for fatal coronary heart disease using Apple Watch ECGs

Background

Fatal coronary heart disease (FCHD) is often described as sudden cardiac death (affects >4 million people/year), where coronary artery disease is the only identified condition. Electrocardiographic artificial intelligence (ECG-AI) models for FCHD risk prediction using ECG data from wearable devices could enable wider screening/monitoring efforts.

Objectives

To develop a single-lead ECG-based deep learning model for FCHD risk prediction and assess concordance between clinical and Apple Watch ECGs.

Methods

An FCHD single-lead ("lead I" from 12-lead ECGs) ECG-AI model was developed using 167,662 ECGs (50,132 patients) from the University of Tennessee Health Sciences Center. Eighty percent of the data (5-fold cross-validation) was used for training and 20% as a holdout. Cox proportional hazards (CPH) models incorporating ECG-AI predictions with age, sex, and race were also developed. The models were tested on paired clinical single-lead and Apple Watch ECGs from 243 St. Jude Lifetime Cohort Study participants. The correlation and concordance of the predictions were assessed using Pearson correlation (R), Spearman correlation (ρ), and Cohen's kappa.

Results

The ECG-AI and CPH models resulted in AUC = 0.76 and 0.79, respectively, on the 20% holdout and AUC = 0.85 and 0.87 on the Atrium Health Wake Forest Baptist external validation data. There was moderate-strong positive correlation between predictions (R = 0.74, ρ = 0.67, and κ = 0.58) when tested on the 243 paired ECGs. The clinical (lead I) and Apple Watch predictions led to the same low/high-risk FCHD classification for 99% of the participants. CPH prediction correlation resulted in an R = 0.81, ρ = 0.76, and κ = 0.78.

Conclusion

Risk of FCHD can be predicted from single-lead ECGs obtained from wearable devices and are statistically concordant with lead I of a 12-lead ECG.

Detection of Arrhythmias Using Smartwatches-A Systematic Literature Review

Smartwatches represent one of the most widely adopted technological innovations among wearable devices. Their evolution has equipped them with an increasing array of features, including the capability to record an electrocardiogram. This functionality allows users to detect potential arrhythmias, enabling prompt intervention or monitoring of existing arrhythmias, such as atrial fibrillation. In our research, we aimed to compile case reports, case series, and cohort studies from the Web of Science, PubMed, Scopus, and Embase databases published until 1 August 2023. The search employed keywords such as "Smart Watch", "Apple Watch", "Samsung Gear", "Samsung Galaxy Watch", "Google Pixel Watch", "Fitbit", "Huawei Watch", "Withings", "Garmin", "Atrial Fibrillation", "Supraventricular Tachycardia", "Cardiac Arrhythmia", "Ventricular Tachycardia", "Atrioventricular Nodal Reentrant Tachycardia", "Atrioventricular Reentrant Tachycardia", "Heart Block", "Atrial Flutter", "Ectopic Atrial Tachycardia", and "Bradyarrhythmia." We obtained a total of 758 results, from which we selected 57 articles, including 33 case reports and case series, as well as 24 cohort studies. Most of the scientific works focused on atrial fibrillation, which is often detected using Apple Watches. Nevertheless, we also included articles investigating arrhythmias with the potential for circulatory collapse without immediate intervention. This systematic literature review provides a comprehensive overview of the current state of research on arrhythmia detection using smartwatches. Through further research, it may be possible to develop a care protocol that integrates arrhythmias recorded by smartwatches, allowing for timely access to appropriate medical care for patients. Additionally, continuous monitoring of existing arrhythmias using smartwatches could facilitate the assessment of the effectiveness of prescribed therapies.

The use of smartwatch electrocardiogram beyond arrhythmia detection

The adoption of wearables in medicine has expanded worldwide with a rapidly growing number of consumers and new features capable of real-time monitoring of health parameters such as the ability to record and transmit a single-lead electrocardiogram (ECG). Smartwatch ECGs are increasingly used but current smartwatches only screen for atrial fibrillation (AF). Most of the literature has focused on analyzing the smartwatch ECG accuracy for the detection of AF or other tachycardias. As with the conventional ECG, this tool may be used for many more purposes than only detection of AF. The objectives of this review are to describe the published literature regarding the accuracy and clinical value of recording a smartwatch ECG in other situations than diagnosis of tachycardia and discuss possible techniques to optimize the diagnostic yield.

The Use of Electrocardiogram Smartwatches in Patients with Cardiac Implantable Electrical Devices

Unlimited access to ECGs using an over-the-counter smartwatch constitutes a real revolution for our discipline, and the application is rapidly expanding to include patients with cardiac implantable electronic devices (CIEDs) such as pacemakers (PMs) and implantable cardioverter defibrillators (ICDs). CIEDs require periodic evaluation and adjustment by healthcare professionals. In addition, implanted patients often present with symptoms that may be related to their PMs or ICDs. An ECG smartwatch could reveal information about device functioning, confirm normal device function, or aid in the case of device troubleshooting. In this review, we delve into the available evidence surrounding smartwatches with ECG registration and their integration into the care of patients with implanted pacemakers and ICDs. We explore safety considerations and the benefits and limitations associated with these wearables, drawing on relevant studies and case series from our own experience. By analyzing the current landscape of this emerging technology, we aim to provide a comprehensive overview that facilitates informed decision-making for both healthcare professionals and patients.

Navigating the Role of Smartwatches in Cardiac Fitness Monitoring: Insights From Physicians and the Evolving Landscape

Alongside the advancement of technology, wearable devices like smartwatches have widely been used for monitoring heartbeat, SpO2, EKG, and pacemaker activity. However, the global question is- can they be as effective as our standard diagnostic tests- electrocardiogram and echocardiography? Reported in the studies, smartwatches to the gold standard Holter monitoring for recognizing irregular pulse showed good sensitivity (98.2%), specificity (98.1%), and accuracy (98.1%). Smartwatches can be good enough for helping people get long-term monitoring of cardiac fitness and early diagnosis of atrial fibrillation but physicians shouldn't completely rely on them and perform standard investigations once the patient with symptoms visits them. We are also concerned that there must be certain rules and regulations for FDA approval of smartwatches to maintain standard criteria before they are released in the market.

Improving Automatic Smartwatch Electrocardiogram Diagnosis of Atrial Fibrillation by Identifying Regularity within Irregularity

Smartwatches equipped with automatic atrial fibrillation (AF) detection through electrocardiogram (ECG) recording are increasingly prevalent. We have recently reported the limitations of the Apple Watch (AW) in correctly diagnosing AF. In this study, we aim to apply a data science approach to a large dataset of smartwatch ECGs in order to deliver an improved algorithm. We included 723 patients (579 patients for algorithm development and 144 patients for validation) who underwent ECG recording with an AW and a 12-lead ECG (21% had AF and 24% had no ECG abnormalities). Similar to the existing algorithm, we first screened for AF by detecting irregularities in ventricular intervals. However, as opposed to the existing algorithm, we included all ECGs (not applying quality or heart rate exclusion criteria) but we excluded ECGs in which we identified regular patterns within the irregular rhythms by screening for interval clusters. This "irregularly irregular" approach resulted in a significant improvement in accuracy compared to the existing AW algorithm (sensitivity of 90% versus 83%, specificity of 92% versus 79%, p < 0.01). Identifying regularity within irregular rhythms is an accurate yet inclusive method to detect AF using a smartwatch ECG.

Accuracy of Artificial Intelligence-Based Technologies for the Diagnosis of Atrial Fibrillation: A Systematic Review and Meta-Analysis

Atrial fibrillation (AF) is the most common arrhythmia with a high burden of morbidity including impaired quality of life and increased risk of thromboembolism. Early detection and management of AF could prevent thromboembolic events. Artificial intelligence (AI)--based methods in healthcare are developing quickly and can be proved as valuable for the detection of atrial fibrillation. In this metanalysis, we aim to review the diagnostic accuracy of AI-based methods for the diagnosis of atrial fibrillation. A predetermined search strategy was applied on four databases, the PubMed on 31 August 2022, the Google Scholar and Cochrane Library on 3 September 2022, and the Embase on 15 October 2022. The identified studies were screened by two independent investigators. Studies assessing the diagnostic accuracy of AI-based devices for the detection of AF in adults against a gold standard were selected. Qualitative and quantitative synthesis to calculate the pooled sensitivity and specificity was performed, and the QUADAS-2 tool was used for the risk of bias and applicability assessment. We screened 14,770 studies, from which 31 were eligible and included. All were diagnostic accuracy studies with case-control or cohort design. The main technologies used were: (a) photoplethysmography (PPG) with pooled sensitivity 95.1% and specificity 96.2%, and (b) single-lead ECG with pooled sensitivity 92.3% and specificity 96.2%. In the PPG group, 0% to 43.2% of the tracings could not be classified using the AI algorithm as AF or not, and in the single-lead ECG group, this figure fluctuated between 0% and 38%. Our analysis showed that AI-based methods for the diagnosis of atrial fibrillation have high sensitivity and specificity for the detection of AF. Further studies should examine whether utilization of these methods could improve clinical outcomes.

Artificial intelligence for detection of ventricular oversensing: Machine learning approaches for noise detection within nonsustained ventricular tachycardia episodes remotely transmitted by pacemakers and implantable cardioverter-defibrillators

BACKGROUND

Pacemakers (PMs) and implantable cardioverter-defibrillators (ICDs) increasingly automatically record and remotely transmit nonsustained ventricular tachycardia (NSVT) episodes, which may reveal ventricular oversensing.

OBJECTIVES

We aimed to develop and validate a machine learning algorithm that accurately classifies NSVT episodes transmitted by PMs and ICDs in order to lighten health care workload burden and improve patient safety.

METHODS

PMs or ICDs (Boston Scientific, St Paul, MN) from 4 French hospitals with ≥1 transmitted NSVT episode were split into 3 subgroups: training set, validation set, and test set. Each NSVT episode was labeled as either physiological or nonphysiological. Four machine learning algorithms-2DTF-CNN, 2D-DenseNet, 2DTF-VGG, and 1D-AgResNet-were developed using training and validation data sets. Accuracies of the classifiers were compared with an analysis of the remote monitoring team of the Bordeaux University Hospital using F2 scores (favoring sensitivity over predictive positive value) using an independent test set.

RESULTS

A total of 807 devices transmitted 10,471 NSVT recordings (82% ICD; 18% PM), of which 87 devices (10.8%) transmitted 544 NSVT recordings with nonphysiological signals. The classification by the remote monitoring team resulted in an F2 score of 0.932 (sensitivity 95%; specificity 99%) The 4 machine learning algorithms showed high and comparable F2 scores (2DTF-CNN: 0.914; 2D-DenseNet: 0.906; 2DTF-VGG: 0.863; 1D-AgResNet: 0.791), and only 1D-AgResNet had significantly different labeling from that of the remote monitoring team.

CONCLUSION

Machine learning algorithms were accurate in detecting nonphysiological signals within electrograms transmitted by PMs and ICDs. An artificial intelligence approach may render remote monitoring less resourceful and improve patient safety.

Accuracy in detecting atrial fibrillation in single-lead ECGs: an online survey comparing the influence of clinical expertise and smart devices

BACKGROUND

Manual interpretation of single-lead ECGs (SL-ECGs) is often required to confirm a diagnosis of atrial fibrillation. However accuracy in detecting atrial fibrillation via SL-ECGs may vary according to clinical expertise and choice of smart device.

AIMS

To compare the accuracy of cardiologists, internal medicine residents and medical students in detecting atrial fibrillation via SL-ECGs from five different smart devices (Apple Watch, Fitbit Sense, KardiaMobile, Samsung Galaxy Watch, Withings ScanWatch). Participants were also asked to assess the quality and readability of SL-ECGs.

METHODS

In this prospective study (BaselWearableStudy, NCT04809922), electronic invitations to participate in an online survey were sent to physicians at major Swiss hospitals and to medical students at Swiss universities. Participants were asked to classify up to 50 SL-ECGs (from ten patients and five devices) into three categories: sinus rhythm, atrial fibrillation or inconclusive. This classification was compared to the diagnosis via a near-simultaneous 12-lead ECG recording interpreted by two independent cardiologists. In addition, participants were asked their preference of each manufacturer's SL-ECG.

RESULTS

Overall, 450 participants interpreted 10,865 SL-ECGs. Sensitivity and specificity for the detection of atrial fibrillation via SL-ECG were 72% and 92% for cardiologists, 68% and 86% for internal medicine residents, 54% and 65% for medical students in year 4-6 and 44% and 58% for medical students in year 1-3; p <0.001. Participants who stated prior experience in interpreting SL-ECGs demonstrated a sensitivity and specificity of 63% and 81% compared to a sensitivity and specificity of 54% and 67% for participants with no prior experience in interpreting SL-ECGs (p <0.001). Of all participants, 107 interpreted all 50 SL-ECGs. Diagnostic accuracy for the first five interpreted SL-ECGs was 60% (IQR 40-80%) and diagnostic accuracy for the last five interpreted SL-ECGs was 80% (IQR 60-90%); p <0.001. No significant difference in the accuracy of atrial fibrillation detection was seen between the five smart devices; p = 0.33. SL-ECGs from the Apple Watch were considered as having the best quality and readability by 203 (45%) and 226 (50%) participants, respectively.

CONCLUSION

SL-ECGs can be challenging to interpret. Accuracy in correctly identifying atrial fibrillation depends on clinical expertise, while the choice of smart device seems to have no impact.

Electrocardiogram Devices for Home Use: Technological and Clinical Scoping Review

BACKGROUND

Electrocardiograms (ECGs) are used by physicians to record, monitor, and diagnose the electrical activity of the heart. Recent technological advances have allowed ECG devices to move out of the clinic and into the home environment. There is a great variety of mobile ECG devices with the capabilities to be used in home environments.

OBJECTIVE

This scoping review aimed to provide a comprehensive overview of the current landscape of mobile ECG devices, including the technology used, intended clinical use, and available clinical evidence.

METHODS

We conducted a scoping review to identify studies concerning mobile ECG devices in the electronic database PubMed. Secondarily, an internet search was performed to identify other ECG devices available in the market. We summarized the devices' technical information and usability characteristics based on manufacturer data such as datasheets and user manuals. For each device, we searched for clinical evidence on the capabilities to record heart disorders by performing individual searches in PubMed and ClinicalTrials.gov, as well as the Food and Drug Administration (FDA) 510(k) Premarket Notification and De Novo databases.

RESULTS

From the PubMed database and internet search, we identified 58 ECG devices with available manufacturer information. Technical characteristics such as shape, number of electrodes, and signal processing influence the capabilities of the devices to record cardiac disorders. Of the 58 devices, only 26 (45%) had clinical evidence available regarding their ability to detect heart disorders such as rhythm disorders, more specifically atrial fibrillation.

CONCLUSIONS

ECG devices available in the market are mainly intended to be used for the detection of arrhythmias. No devices are intended to be used for the detection of other cardiac disorders. Technical and design characteristics influence the intended use of the devices and use environments. For mobile ECG devices to be intended to detect other cardiac disorders, challenges regarding signal processing and sensor characteristics should be solved to increase their detection capabilities. Devices recently released include the use of other sensors on ECG devices to increase their detection capabilities.

Using a Smartwatch to Record Precordial Electrocardiograms: A Validation Study

Smartwatches that support the recording of a single-lead electrocardiogram (ECG) are increasingly being used beyond the wrist, by placement on the ankle and on the chest. However, the reliability of frontal and precordial ECGs other than lead I is unknown. This clinical validation study assessed the reliability of an Apple Watch (AW) to obtain conventional frontal and precordial leads as compared to standard 12-lead ECGs in both subjects without known cardiac anomalies and patients with underlying heart disease. In 200 subjects (67% with ECG anomalies), a standard 12-lead ECG was performed, followed by AW recordings of the standard Einthoven leads (leads I, II, and III) and precordial leads V1, V3, and V6. Seven parameters (P, QRS, ST, and T-wave amplitudes, PR, QRS, and QT intervals) were compared through a Bland-Altman analysis, including the bias, absolute offset, and 95% limits of agreement. AW-ECGs recorded on the wrist but also beyond the wrist had similar durations and amplitudes compared to standard 12-lead ECGs. Significantly greater amplitudes were measured by the AW for R-waves in precordial leads V1, V3, and V6 (+0.094 mV, +0.149 mV, +0.129 mV, respectively, all p < 0.001), indicating a positive bias for the AW. AW can be used to record frontal, and precordial ECG leads, paving the way for broader clinical applications.

Feasibility and Diagnostic Value of Recording Smartwatch Electrocardiograms in Neonates and Children

OBJECTIVE

The objective of this study was to evaluate the agreement of smartwatch-derived single-lead electrocardiogram (ECG) recordings with 12-lead ECGs for diagnosing electrocardiographic abnormalities.

STUDY DESIGN

A 12-lead ECG and an ECG using Apple Watch were obtained in 110 children (aged 1 week to 16 years) with normal (n = 75) or abnormal (n = 35) 12-lead ECGs (atrioventricular block [7], supraventricular tachycardia [SVT] {5}, bundle branch block [12], ventricular preexcitation [6], long QT [5]). In children aged <6 years, the ECG recording was performed with the active participation of an adult who applied the neonate or child's finger to the crown of the watch. In older children, tracings were obtained after brief teaching without adult guidance. All 12-lead ECGs were independently evaluated by 2 blinded cardiologists. Apple Watch ECGs were independently evaluated by another blinded cardiologist.

RESULTS

In 109 children (99.1%), the smartwatch tracing was of sufficient quality for evaluation. Smartwatch tracings were 84% sensitive and 100% specific for the detection of an abnormal ECG. All 75 normal tracings were correctly identified. Of the 35 children with abnormalities on 12-lead ECGs, 5 (14%) were missed, most often because of baseline wander and artifacts. Rhythm disorders (atrioventricular block or SVT) and bundle branch blocks were correctly detected in most cases (11 of 12 and 11 of 12, respectively); preexcitation and long QT was detected in 4 of 6 and 4 of 5, respectively.

CONCLUSION

Smartwatch ECGs recorded with parental assistance in children aged up to 6 years and independently in older children have the potential to detect clinically relevant conditions.

An implantable loop recorder or smartphone based single-lead electrocardiogram to detect arrhythmia in adults with congenital heart disease?

Background

The European Society of Cardiology (ESC) guidelines for the management of adult congenital heart disease (ACHD) recommend screening in patients at risk for arrhythmic events. However, the optimal mode of detection is unknown.

Methods

Baseline and follow-up data of symptomatic ACHD patients who received an implantable loop recorder (ILR) or who participated in a smartphone based single-lead electrocardiogram study were collected. The primary endpoint was time to first detected arrhythmia.

Results

In total 116 ACHD patients (mean age 42 years, 44% male) were studied. The ILR group (n = 23) differed from the smartphone based single-lead electrocardiogram group (n = 93) in having a greater part of males and had more severe CHD and (near) syncope as qualifying diagnosis. In the smartphone based single-lead electrocardiogram group history of arrhythmia and palpitations were more frequent (all p < 0.05). Monitoring was performed for 40 and 79 patient-years for the ILR- and smartphone based single-lead electrocardiogram group, respectively. Arrhythmias occurred in 33 patients with an equal median time for both groups to first arrhythmia of 3 months (HR of 0.7, p = 0.81). Furthermore, atrial fibrillation occurred most often (n = 16) and common therapy changes included medication changes (n = 7) and implantation of pacemaker or Implantable Cardioverter Defibrillator (ICD) (N = 4). Symptoms or mode of detection were not a determinant of the first event.

Conclusion

Non-invasive smartphone based single-lead electrocardiogram monitoring could be an acceptable alternative for ILR implantation in detecting arrhythmia in symptomatic ACHD patients in respect to diagnostic yield, safety and management decisions, especially in those without syncope.

Analysis of the amplified p-wave enables identification of patients with atrial fibrillation during sinus rhythm

Aim

This study sought to develop and validate diagnostic models to identify individuals with atrial fibrillation (AF) using amplified sinus-p-wave analysis.

Methods

A total of 1,492 patients (491 healthy controls, 499 with paroxysmal AF and 502 with persistent AF) underwent digital 12-lead-ECG recording during sinus rhythm. The patient cohort was divided into training and validation set in a 3:2 ratio. P-wave indices (PWI) including duration of standard p-wave (standard PWD; scale at 10 mm/mV, sweep speed at 25 mm/s) and amplified sinus-p-wave (APWD, scale at 60-120 mm/mV, sweep speed at 100 mm/s) and advanced inter-atrial block (aIAB) along with other clinical parameters were used to develop diagnostic models using logistic regression. Each model was developed from the training set and further tested in both training and validation sets for its diagnostic performance in identifying individuals with AF.

Results

Compared to standard PWD (Reference model), which achieved an AUC of 0.637 and 0.632, for training and validation set, respectively, APWD (Basic model) importantly improved the accuracy to identify individuals with AF (AUC = 0.86 and 0.866). The PWI-based model combining APWD, aIAB and body surface area (BSA) further improved the diagnostic performance for AF (AUC = 0.892 and 0.885). The integrated model, which further combined left atrial diameter (LAD) with parameters of the PWI-based model, achieved optimal diagnostic performance (AUC = 0.916 and 0.902).

Conclusion

Analysis of amplified p-wave during sinus rhythm allows identification of individuals with atrial fibrillation.

Accuracy and clinical relevance of the single-lead Apple Watch electrocardiogram to identify atrial fibrillation

Background

The Apple Watch (AW) is the first commercially available wearable device with built-in electrocardiogram (ECG) electrodes to perform a single-lead ECG to detect atrial fibrillation (AF).

Methods

Patients with AF who were scheduled for electrical cardioversion (ECV) were included in this study. The AW ECGs were obtained pre-ECV and post-ECV. In case of an unclassified recording, the AW ECG was obtained up to 3 times. The 12-lead ECG was used as the reference standard. Sensitivity, specificity, and kappa coefficient were calculated.

Results

In total, 74 patients were included. Mean age was 67.1 ± 12.3 years and 20.3% were female. In total 65 AF and 64 sinus rhythm measurements were obtained. The first measurement with the AW showed a sensitivity of 93.5% and specificity of 100% (κ = 0.94). A second measurement resulted in a sensitivity of 94.6% and specificity of 100% (κ = 0.95). A third measurement resulted in a sensitivity of 93% and a specificity of 96.5% (κ = 0.90). Adjudication of unclassified recordings by a physician reduced the total unclassified recordings from 27.9% to 1.6%, but also reduced the accuracy. The kappa coefficient for unclassified single-lead ECGs was 0.58.

Conclusion

The single-lead ECG of the AW shows a high accuracy for identifying AF in a clinical setting. Repeating the recording once decreases the total of unclassified recordings; however, a third recording resulted in a lower accuracy and the occurrence of false-positive measurements. Unclassified results of the AW can be reduced by physicians' interpretation of the single-lead ECG; however, the interrater agreement is only moderate.

Six-lead device superior to single-lead smartwatch ECG in atrial fibrillation detection

This was a head-to-head comparative study on different electrocardiogram (ECG)-based smartwatches and devices for atrial fibrillation detection. We prospectively included 220 patients scheduled for electrical cardioversion and recorded ECGs with 3 different devices (Withings Move ECG, Apple Watch 5, Kardia Mobile 6-leads) as well as the standard 12-lead ECG (gold standard), both before and after cardioversion. All atrial fibrillation detection algorithms had high accuracy (sensitivity and specificity: 91-99%) but were hampered by uninterpretable recordings (20-24%). In cardiologists' interpretation, the 6-lead device was superior (sensitivity 99%, specificity 97%) to both single-lead smartwatches (P < .05) for atrial fibrillation detection.