Detection Of Atrial Fibrillation Research Articles

MRMEBP: a unified framework for online detection of atrial fibrillation utilizing deep learning

mRMEBP: a unified framework for online detection of atrial fibrillation utilizing deep learning

ECG-based machine learning model for AF identification in patients with first ischemic stroke.

The recurrence rate of strokes associated with atrial fibrillation (AF) can be substantially reduced through the administration of oral anticoagulants. However, previous studies have not demonstrated a clear benefit from the universal application of oral anticoagulants in patients with embolic stroke of undetermined source. Timely detection of AF remains a challenge in patients with stroke. This study aims to develop a convolutional neural network (CNN) model to accurately identify patients with AF using a 12-lead sinus-rhythm electrocardiogram (ECG) recorded around the time of the first ischemic stroke. Additionally, this study also evaluates the model's ability to predict future occurrence of AF. A CNN model was trained with ECG data from patients at Taipei Veterans General Hospital. External validation was performed on ischemic stroke patients from National Taiwan University Hospital. The model's performance was assessed for detecting AF at the stroke event and predicting future AF occurrences. The model demonstrated an area under curve (AUC) of 0.91 for internal validation and 0.69 for external validation in identifying AF at the stroke event, with sensitivity and negative predictive value both achieving 97%. Kaplan-Meier survival analysis of patients without a prior diagnosis of AF revealed a significant increase in future AF incidence among the high-risk group identified by the model (adjusted hazard ratio: 4.06; 95% confidence interval: 2.74-6.00). The CNN model effectively identifies AF in stroke patients using 12-lead ECGs and predicts future AF events, facilitating early anticoagulation therapy and potentially reducing recurrent stroke risk. Further prospective studies are warranted to confirm these findings.

Screening for Atrial Fibrillation During Routine Automated Blood Pressure Measurement in General Population Aged 65 Years and Above: EMENO National Epidemiological Survey in Greece.

Atrial fibrillation (AF) is common with advanced aging and is often asymptomatic and undiagnosed. A blood pressure (BP) monitor with embedded algorithm for detecting AF with high diagnostic accuracy during routine automated BP measurement was used to screen individuals aged ≥ 65 years. The EMENO health survey (2013-2016) applied a multi-stage stratified random sampling method to select a random general adult population sample in Greece. Participants aged ≥ 65 years were screened for AF using an automated oscillometric device with AF detecting algorithm (Microlife BPA100 Plus AFIB) with triplicate measurements on a single occasion. A total of 975 individuals with AF data on at least one BP measurement were analysed [(mean (SD) age 73.2 (6.6) years, men 45%, hypertensives 79.3%). AF was detected in all 3 BP readings in 65 individuals (6.3%, AF group), in none of 3 readings in 661 (73.6%, non-AF group), and in 1-2 readings of 3 or less readings in 249 (20.1%, uncertain AF group). AF group individuals were older (p < 0.001), had lower systolic BP (p < 0.05) and more frequent cardiovascular disease (p < 0.01) compared to the non-AF and uncertain AF group. Moreover, 44.6% were unaware of their AF and all required anticoagulant therapy (CHA2DS2-VASc ≥ 1/≥2, men/women). About 6% of individuals aged ≥ 65 years in Greece appeared to have AF, and 44.6% of them were unaware and requiring anticoagulant treatment. AF screening during routine automated BP measurement appears to be a useful tool for early detection of asymptomatic AF in individuals aged ≥ 65 years.

Accuracy of the Apple Watch in Detecting Atrial Fibrillation Among Patients Undergoing 24-Hour Holter Monitoring: A Prospective, Pragmatic Study

BACKGROUND: As smartwatches with atrial fibrillation detection features gain popularity, it is important to assess the accuracy of these devices to guide decision-making. OBJECTIVES: Our study aimed to assess the sensitivity and specificity of the irregular rhythm notification and the electrocardiogram (ECG)–based detection features of a commonly used smart wearable device (Apple Watch) in detecting atrial fibrillation. METHODS: This was a prospective, pragmatic study conducted in Perpetual Succour Hospital–Cebu Heart Institute from August 2023 to January 2024. To assess the irregular rhythm notification feature, participants were asked to wear an Apple Watch alongside a 24-hour Holter monitor to verify notifications. For the ECG-based detection feature, participants had to tap the crown of the Apple Watch for 30 seconds to get a single-lead ECG similar to a lead I ECG tracing. They were instructed to get manual ECGs hourly, or more often while awake. Irregular rhythm notifications and ECG readings were then compared with that of the 24-hour Holter monitor. Sensitivity and specificity were then computed. RESULTS: A total of 140 participants consented to join after full study disclosure. The irregular rhythm notification feature of the Apple Watch exhibited a low sensitivity of 21.4% but achieved a high specificity of 100% in detecting atrial fibrillation. Meanwhile, the ECG-based detection feature, analyzed from 1295 manually taken ECGs with interpretable sinus rhythm or atrial fibrillation, demonstrated a high level of agreement with the Holter monitor, with a sensitivity of 100% and a specificity of 99.1%. CONCLUSION: The low sensitivity of the irregular rhythm notification feature of the Apple Watch in detecting atrial fibrillation cautions against relying on it as a primary screening tool. However, the high concordance of manually taken Apple Watch ECGs positions the device as a robust tool for detecting atrial fibrillation through manual ECG detection. KEYWORDS: Apple Watch, atrial fibrillation, smartwatches

Consumer Wearables-Advancing Atrial Fibrillation Care or Too Much Information?

This Viewpoint describes the diffusion of direct-to-consumer wearable devices capable of early atrial fibrillation detection and the unclear implications for initiation of anticoagulation therapy based on such findings.

Continuous heart rate monitoring in atrial fibrillation with multispectral photoplethysmography

Heart attacks and strokes account for more than four out of every five Cardiovascular Disease (CVD) deaths, with one-third occurring before age 70 years. Atrial fibrillation (AF), a common cause of ischemic strokes, can significantly reduce mortality rates through continuous monitoring, timely diagnosis, accurate confirmation, and early prognosis. This can be achieved by employing Multi-Spectral Photoplethysmography (MSPPG) technology as an alternative diagnostic tool to Electrocardiography (ECG). This investigation is implemented by determining the Heart Rate (HR) of AF-affected patients using a developed MSPPG-based wrist-worn device, keeping ECG as a standard gold reference. The Multispectral-based wrist-worn wearable Photoplethysmography device employs an optical bio-sensing, integrated BIOFY sensor to acquire the MSPPG signals using two green (526 nm), one red (660 nm) and one infrared (950 nm) LED and two photodetectors (Broadband detector, IR-cut detector). The MSPPG signals acquired using a microcontroller unit are stored in an On-device storage system simultaneously in the IoT Platform. Statistical analysis techniques were employed to compare the HR detected using the MSPPG technique with the HR measured from the gold-standard ECG. The accuracy paves the way for implementing the proposed MSPPG technology in detecting AF conditions in terms of HR, further laying the foundation for earlier detection of AF. Wearable PPG-based devices face a significant issue of motion artifacts that can be overcome using the MSPPG technique. The research advancements can potentially increase the accessibility of HR monitoring in AF patients, conveniently integrating into daily routines without restricting day-to-day activities.

Electrocardiogram sampling frequency for optimal performance of complexity analysis and machine learning models: Discrimination between subjects with and without paroxysmal atrial fibrillation using sinus rhythm electrocardiograms

BackgroundCurrent clinical practice to diagnose atrial fibrillation (AF) requires repeated, episodic monitoring and significantly underperform in their ability to detect AF episodes. ObjectiveThere is therefore potential for artificial intelligence based methods to assist in the detection of AF. Better understanding the optimal parameters for this detection can potentially improve the sensitivity in detecting AF. MethodTen second, 12-lead electrocardiogram signals were analysed using complexity algorithms combined with machine learning techniques to predict patients who had a previously detected AF episode but had since returned to normal sinus rhythm. An investigation was performed into the impact of sampling frequency of the electrocardiogram signal on the accuracy of the machine learning models used. ResultsUsing a single complexity algorithm showed a peak accuracy of 0.69 when using signals sampled at 125Hz. In particular it was noted that improved accuracy occurred when using lead V6 compared to other available leads. ConclusionBased on these results there is potential for 12-lead electrocardiogram signals to be recorded at 125Hz as standard and used in conjunction with complexity analysis to aid in the detection of subjects with AF.

MOBILE HEALTH: ECG MONITORING FOR ATRIAL FIBRILLATION DETECTION

This systematic review examines the use of mobile health (mHealth) technologies, such as smartphone apps and wearable devices, for detecting atrial fibrillation (AF). The review includes studies assessing the quality and performance of mobile apps and smartwatch algorithms, as well as the clinical validity of ECG measurement and photoplethysmography technology. Findings suggest diverse results, with some studies indicating promising sensitivity and specificity for AF detection. However, there is a need for further research to develop reliable algorithms for diagnosing and managing cardiovascular conditions using mHealth technologies.

Accuracy of Smartwatches for AtrialFibrillation Detection Over Time: Insights From the Basel Wearable Study.

Accuracy of Smartwatches for AtrialFibrillation Detection Over Time: Insights From the Basel Wearable Study.

How can the detection of AF be improved? Insights from patient journeys and primary care clinicians

Abstract Background Atrial fibrillation (AF) is underdiagnosed and it is estimated that a third of people with AF remain undiagnosed. AF is complex to detect because it can be intermittent and some patients are asymptomatic. When symptoms are present they can be non-specific. In the absence of a screening programme to detect AF it is important to identify ways to improve its detection of AF in order to reduce the risk of AF-related stroke. Purpose We conducted a qualitative study in primary care to explore the patient path to diagnosis of AF and identify ways to improve detection Methods We recruited patients with a recent diagnosis of AF (&amp;lt; 6 months) and interviewed them to understand their path to diagnosis. We interviewed GPs and practice nurses to explore their experience of detecting AF in primary care. The interviews were audio-recorded and transcribed, and the data were analysed using framework analysis. Results Thirty patients and ten primary care clinicians were interviewed. The presence and nature of symptoms experienced by patients and how they interpreted these were significant in the pathways to diagnosis. Patients with acute symptoms presented to A &amp;E and had a quick diagnosis while asymptomatic patients were diagnosed incidentally. Patients with non-specific symptoms generally did not perceive the symptoms as serious, and many delayed seeing a healthcare professional. Their experiences in primary care aligned with findings from interviews with primary care clinicians, who acknowledged AF may not necessarily be the initial suspicion when a patient presents with certain non-specific symptoms. Primary care clinicians described narratives of good practice in the form of opportunistic pulse palpation, as well as challenges of detecting AF in primary care. These included lack of access to 24-hour ambulatory ECG in primary care and limited opportunities to detect AF due to healthcare assistants taking on more responsibilities. Conclusions A number of opportunities to improve the detection of AF can be drawn from these findings. Public education about AF, its signs and symptoms would increase patient knowledge and presentation to healthcare professionals. Standardizing good practice of opportunistic pulse palpation in primary care and training healthcare assistants to perform this is also indicated. Access to 24-hour ambulatory ECG in primary care will reduce delays in diagnosis. Further research on which patients are likely to have undiagnosed AF will facilitate detection in primary care.

Atrial fibrillation detection on implantable loop recorders in embolic stroke of undetermined source

Abstract Introduction Embolic stroke of undetermined source (ESUS) represents 17% of all ischaemic strokes, and has a 5% annual stroke recurrence risk. Recent large randomised control trials failed to demonstrate a benefit of empirical anticoagulation for all ESUS patients, with sub-analyses supporting the need for a tailored approach to those patients at greatest risk of atrial fibrillation (AF) development. Understanding how the increased incidence of ILR-detected AF in the ESUS cohort, compares with ILR-detected AF in other cohorts is an essential step in understanding the significance of this finding. Purpose To compare the incidence of ILR-detected AF following ESUS against patients undergoing ILR implantation for non-stroke indications for example syncope and palpitations. Methods We retrospectively studied all patients without known AF who were referred to our institution for ILR implantation following ESUS, syncope or palpitations from 2009 to 2019. The primary endpoint was any detection of AF or atrial flutter (AFL) by ILR. Kaplan-Meier and multivariable Cox-regression analyses were utilised to account for time-to-event. Results A total of 750 patients were included and followed up for a mean of 731 days (SD 443). An ILR was implanted following ESUS in 323 patients and for another reason in 427 patients (figure 1). The incidence of AF was significantly higher among patients with ESUS compared to the non-ESUS group; 48.6% versus 13.8% (for any duration of AF) and 32.2% versus 12.4% (for AF lasting &amp;gt; 30 seconds) both p &amp;lt;0.001. Kaplan-Meier analysis (figure 2), showed significantly higher incidence of AF for ESUS for any duration of AF. This was driven predominantly by AF duration &amp;lt;6 minutes and between 5.5 hours and 24 hours. Importantly, in a multivariable Cox regression analysis, ESUS independently conferred an almost 5-fold increase in the hazard for any duration of AF (Hazard Ratio 4.948, 95% confidence interval 3.587 – 6.825, p&amp;lt;0.001). Conclusion The incidence of AF is significantly higher amongst ESUS versus non-ESUS patients monitored constantly by an ILR. A high number of ESUS survivors have short duration AF episodes. Further work is needed to determine the optimal treatment strategy of these AF episodes in ESUS considering AF duration and overall burden.

An explainable AI for trustworthy detection of atrial fibrillation on reduced lead ECGs in mobile applications

Abstract Introduction Artificial intelligence (AI), particularly deep learning (DL), has demonstrated high performance in various diagnostic problems, including the detection of paroxysmal atrial fibrillation (AF) from electrocardiograms (ECGs). In mobile applications, these approaches can help to detect AF at an early stage and thus prevent possible secondary diseases. However, due to their black-box nature, DL approaches lack explainability and interpretability. Furthermore, 12-lead ECGs are the clinical standard, but mobile devices usually only provide 1–3 leads, which are similar but not identical to Einthoven leads and differ in morphology. Purpose We apply a recently introduced explainable ECG analysis architecture (xECGArch) to single-lead ECGs to verify its performance in mobile applications. xECGArch can learn rhythmic and morphological characteristics using two parallel convolutional neural networks of different dimensions (Fig. 1). When combined with methods from explainable AI, the classification can be traced back to rhythmic and morphological characteristics. Methods We used a subset of 9 854 ECGs (n(AF) = n(non-AF) = 4 927) from 4 public databases (Chapman-Shaoxing, CPSC2018, Georgia, and PTB-XL). Noise was removed using a cascade of high-pass filtering at 0.3 Hz and discrete wavelet transform. For each lead, we trained xECGArch on single-lead ECGs of 10 s length to classify AF vs. non-AF using 90% of the ECGs with 5-fold cross-validation and tested it on the remaining randomly selected ECGs (n(AF) = 507, n(non-AF) = 479). Consistent data splitting was ensured for all leads. Model explanations were generated using deep Taylor decomposition. Results Both models achieved minimum 91.1% accuracy for each ECG lead (Fig. 1). On lead I, the morphology model outperformed the rhythm model with 92.8% vs. 91.5% accuracy, while the rhythm model slightly outperformed the morphology model on leads II and III with 94.2%–95.0% accuracy vs. 91.1%–94.0%. Both models achieved maximum accuracy of 95.0% (rhythm) and 94.0% (morphology) on lead II. By combining both models, xECGArch reached 93.7%–95.1% accuracy. The model explanations confirm that the rhythm model primarily considers QRS complexes as relevant, while the morphology model focuses on fibrillatory waves (Fig. 1). Conclusions Both rhythm and morphology models achieved reliable results across all leads, competing with the state-of-the-art for automatic AF detection from single-lead ECGs. The combination of both models in xECGArch further increased the accuracy. xECGArch is based on the medical reading of ECGs by considering rhythm and morphology characteristics and is therefore interpretable by design. The model explanations are consistent with the diagnostic criteria for AF (Fibrillatory waves and absolute arrhythmia) across leads. With high accuracy on different single-lead ECGs and improved interpretability, xECGArch provides a trustworthy method for ECG analysis in mobile applications.

Validation of an artificial intelligence system for holter ECG analysis

Abstract Introduction Holter ECG is a commonly used tool to identify heart rhythm disorders. As physician-based analysis of such Holter ECG is time consuming, various approaches using computerized interpretation have been developed. However, accuracy of these tools remains inferior to physician-based analysis. Thus, this study assesses the diagnostic performance of an artificial intelligence (AI)-powered Holter ECG service in comparison to results from reference databases as well as with physician-based evaluation. Methods 231 ECG recordings, covering 2,261 hours of ECG with annotations for 10,024,315 heartbeats from the following reference databases were used: MIT-BIH Arrhythmia (MITDB), American Heart Association ECG (AHDB), MIT-BIH Atrial Fibrillation (AFDB), Long Term Atrial Fibrillation (LTAFDB). Those recordings were originally collected from a mixed population of inpatients and outpatients between 1975 and 2008. Analysis of these recordings regarding the detection of QRS complexes, QRS classification and detection of Atrial Fibrillation (AF) was performed using a novel service for AI-based ECG analysis. Results were compared to the diagnosis stated within the database. QRS complex detection accuracy was performed on recordings from MITDB and AHADB by checking if time positions are within 150 ms to reference annotations. The QRS classification in three classes (normal, supraventricular, ventricular) was then compared for each detected heartbeat. Mismatches were adjudicated manually by three experts. Episodes of Atrial Fibrillation were detected in recordings from MITDB, AFDB and LTAFDB and compared to the reference episodes from the databases. The accuracy was evaluated based on the overlap with the reference annotations in the time-domain. Results Sensitivity and Precision of heartbeat detection compared to the reference databases exceeded 99% Figure 1 displays the results of heartbeat classification. Normal heartbeats were classified with a sensitivity of 99.6%, precision of 99.9% and specificity of 99.6%. For ventricular heartbeats, a sensitivity of 99.58 %, precision of 99.88 and specificity of 99.99% were achieved. Premature supraventricular contractions were recognized with a sensitivity of 99.58 %, precision of 100% and specificity of 100%. When comparing the 248,187 QRS complex classifications of MITDB and AHADB to the AI-powered analysis, 128 discrepancies were revealed. 43 were due to a discrepancy of convention as these beats were in sections that were also marked as AF episodes by the databases and the AI-powered analysis marks all beats in AF episodes as normal. Of the 85 remaining discrepancies, more than 87% were adjudicated in favor of the AI-powered analysis by the experts. Figure 2 shows the comparison of AF detection results showing Sensitivity and Precision above 99%. Conclusion AI-powered Holter ECG analysis compared to reference ECG stripes exceed 99% of sensitivity, specificity and precision.QRS complex detectionAF/ AFib detection

Quality of life after screening for atrial fibrillation: a substudy from a randomised screening trial

Abstract Background Screening for atrial fibrillation (AF) is rising, yet trials have failed to show efficacy for stroke prevention from screening and subsequent anticoagulation. Quality of life is an important outcome for patients and physicians, however, it is largely unknown how this outcome is affected by screening for AF. The adverse effects of screening should be considered before implementing systematic screening. Purpose To explore changes in quality of life associated with screening for AF and subsequent anticoagulant treatment upon AF detection. Methods We used quality of life data from 6,004 persons with stroke risk factors randomised to usual care (n=4,503) or screening for AF using implantable loop recorder and anticoagulation for AF episodes lasting ≥6 minutes (n=1,501). The validated EQ-5D-5L tool evaluates health-related quality of life in five domains (mobility, selfcare, usual activities, pain/discomfort and anxiety/depression), rendering a final score in EQ-index (worst=-0.76 best=1.00) combined with an EQ-VAS score (worst=0, best=100). EQ-index and EQ-VAS were analysed with linear mixed models from baseline through year three. The domains were also analysed with logistic regression to report the risk of major problems (defined as a rating of 3, 4, or 5 in the range [1-5], where 1 corresponds to no problem and 5 corresponds to an extreme problem within a given domain) overall and separately. Sensitivity analyses with imputation for missing data were performed. Results Data were available for 5,162 (86.0%) participants at baseline and in year three, and 227 (5.0%) in usual care and 65 (4.3%) in the screened group had died before the last assessment. At baseline, 1,202 of 5,162 (23.3%) participants reported major pain/discomfort problems, and 666 of 5,162 (12.9%) reported major mobility problems (Figure 1). Major selfcare problems were the least frequent problem reported by 83 of 5,162 participants (1.6%). The occurrence of major problems was increased after 3 years in all domains and in both randomisation groups, with a 2% higher occurrence of major pain/discomfort in the control group compared to the screened group. At baseline, the average EQ-index was 0.88 (±0.16) and EQ-VAS 78.4 (±16.2). After three years, the EQ-index decreased by -0.05 (-0.05; -0.04), p&amp;lt;0.0001 in the control group vs. -0.04 (-0.05; -0.02) in the screened group and EQ-VAS with -6.06 (-6.54; -5.57) in the control group vs. -5.18 (-6.52; -3.78) in the screened group, resulting in no statistically significant difference between the groups (p=0.063 and p=0.056) (Figure 2). The screened group was equally likely to report at least one major problem in quality of life after three years (odds ratio 0.93 (0.77;1.13)), but was less likely to report major pain/discomfort (odds ratio 0.84 (0.72;0.98)), compared with the control group. Conclusion Systematic and intensive screening for atrial fibrillation did not improve quality of life compared to usual practice.

Screening for Atrial Fibrillation (AF) by handheld single-lead ECG and risk of all-cause mortality stratified by non-AF cardiovascular disease

Abstract Purpose Screening for Atrial fibrillation (AF) has been suggested as a strategy to increase early detection of silent AF. We aimed to evaluate all-cause mortality between patients with screen-detected and clinically-diagnosed AF, and the effect of oral anticoagulation therapy (OAC) on mortality. Methods We prospectively enrolled and performed AF screening using handheld single-lead ECG(AliveCor) on consecutive patients who attended medical outpatient clinics during December 2014 and December 2017. Repeat screening was performed among patients who had &amp;gt;1 clinic visits. All participants were divided into 5 cohorts: (i) previously known AF at baseline (kAF); (ii) initial screen-detected AF (S1-AF); (iii) subsequent screen-detected AF (FU-sAF); (iv) clinically-diagnosed AF during follow-up (FU-cAF); and (v) no AF. Risk of all-cause mortality was estimated by adjusted hazard ratio (HR) using COX proportional hazards regression, adjusting for age, gender, co-morbidities, concurrent chronic use of medications including OAC, with AF detection or clinical diagnosis and OAC therapy handled as time-varying variable. Subgroup analysis was conducted to identify heterogeneity of effect across with or without non-AF cardiovascular disease (CVD). Results Of 11,972 subjects (mean age 76.6±7.8, female 49.2%), 18.7% (n=2,236) had previously kAF, 1.9% (n=223) S1-AF, 0.6% (n=71) FU-sAF, 7.3% (n=867) FU-cAF, and 71.6% (n=8,559) no AF. During a median FU duration of 6.8(IQR: 5.2-8.1) years, patients with kAF (HR=1.8(1.6-2.0)), S1-AF (HR=1.6(1.3-1.9)) and FU-cAF (HR=1.6(1.4-1.8)) were at higher risk of all-cause mortality, compared to individuals without AF, whereas FU-sAF (HR=1.0(0.7-1.4)) had similar low risk of death as no AF. Heterogeneity of effect was detected in subgroup analysis. Among patients without non-AF CVD, FU-sAF (HR=6.0(1.8-19.8)) and FU-cAF (HR=4.7(3.3-6.7)) were at markedly higher risk of all-cause mortality. DOAC (HR=0.57(0.38-0.87)) but not VKA (HR=0.72(0.43-1.15) reduced mortality. However, in patients with CVD other than AF, risk of FU-sAF (HR=1.1(0.8-1.6)) decreased to as low as individuals without AF (P for interaction =0.0039).Both DOAC (HR=0.51(0.46-0.58)) and VKA (HR=0.87(0.78-0.98) were beneficial. Conclusion Likely delayed AF detection suggested by groups of FU-cAF and FU-sAF may be associated with significant rise in risk of mortality among patients without non-AF CVD. The effect diminished when non-AF CVD as potential competing risk existed. The role of VKA in reducing all-cause mortality warrants more studies.

Predicting 1-year atrial fibrillation risk from a patch-based ambulatory ECG monitoring without atrial fibrillation

Abstract Introduction Early identification of atrial fibrillation (AF) can enable interventions that reduce cardiovascular events. However, AF remains undiagnosed in about 10-20% of affected individuals, and paroxysmal AF often remains undetected despite the use of in-clinic ECGs, consumer smartwatch ECGs, or ambulatory ECG monitoring. We therefore investigated whether ambulatory ECG patch-based monitoring without AF can predict AF in the subsequent 1-year post monitoring. Purpose This study aims to validate a machine-learning model for AF prediction within 1-year post-wear, based on the observation of non-AF single-lead ECG recordings. Methods We considered a clinically monitored cohort of 67,296 individuals wearing multiple single-lead ECG multi-day ambulatory monitoring patches as part of real-world care. For all selected individuals, the first patch recorded no occurrences of AF and was worn for a minimum of 6.5 days. We assess the risk of future AF based on the raw ECG, heartrate variability, frequency of ectopic beats, age and gender (ECG model), which is then compared with the risk assessment by a model based on age and gender (Baseline model). We then observe AF incidence on subsequent patches worn within one year from the first patch. We evaluated the model's accuracy by observing the number of AF cases identified (or missed) among the individuals with highest risk according to the model. Results In the clinically monitored cohort, a total of 3463 individuals experienced AF within the next year (incidence 5.1%) after a first AF-free monitoring period of a minimum 6.5 days. The receiver operating characteristic (ROC) curve shows an area under the curve (AUC) for the ECG model of 0.75 (95% CI: [0.74, 0.76]), and for the Baseline model of 0.66 (CI: [0.65, 0.67]). When selecting the top 20% (13,459) participants with the highest risk and verifying who developed AF in the following 1-year, we obtain positive prediction value (PPV) or precision of 12% and sensitivity (recall) of 46% for the ECG model (which correctly identifies 1,594 individuals with incident AF), and of 8% and 29%, respectively, for the Baseline model (p-value&amp;lt;0.001). (Figure) Conclusion A machine learning model using as input an AF-free single-lead ECG can quantify the risk of observing AF in a population including individuals with low burdens of paroxysmal or new onset of AF within 1 year. Further research and validation are needed to enhance the interpretability and clinical applicability of the model, while prediction may be further improved with clinical AF diagnoses ascertained from other data sources. The findings of this study have important implications for the early detection and management of AF, potentially enabling earlier intervention and improved outcomes in individuals with undiagnosed AF.

Validation of diagnostic wearable ECG data from high-risk patients and integration into the atrial fibrillation catheter ablation pathway

Abstract Background Wearables like the Apple Watch (AW) offer non-invasive, long-term monitoring and patient-led, single-lead ECG recordings. Although validated for Atrial Fibrillation (AF) detection in the general population, the accuracy and clinical utility of the single-lead ECG recording feature for the diagnosis of AF has not been evaluated in patients at high risk of atrial arrhythmias.(1) An evaluation of the accuracy of the Wearable-derived labels is important to determine the feasibility of integrating AW-based remote monitoring into clinical pathways. Objective To evaluate the accuracy and data burden from AW-based, patient-led ECG monitoring in patients undergoing AF catheter ablation. Methods Patients referred for first-time AF catheter ablation to the Study Institution between January 2022 and March 2023 were invited to enrol in the AF Follow-up with Apple Watch (AFFU-AW) Trial. Patients randomised to the active arm of the study received 12 months of post-ablation follow-up using a loaned Apple Watch. Participants underwent an education session and were asked to perform daily ECGs; sending through any labelled as AF, inconclusive, or in the context of symptoms. The accuracy of the automated ECG labelling was determined against blinded, independent labelling by two experienced Electrophysiologists. Results 1093 remote transmissions were emailed in for evaluation from 52 patients to the Study team. This was a median of 6 [2, 17] per patient and a receiver burden of 2 [1, 4] ECGs per day. 475 were categorised by the AW as AF with a sensitivity of 0.71 and specificity of 0.96 compared to expert labelling. 342 (31%) were not categorised by the wearable, with an ‘inconclusive’ interpretation (161 [47%]) or ‘high heart rate’ (108 [32%]) as the most cited reasons. 178 (55%) of the uncategorised ECGs were subsequently overread as AF. Accompanying symptoms were reported with 324 (30%) ECGs, most commonly palpitations. Conclusion This is the first reported evaluation of the AF diagnostic feature based on AW ECG recordings as part of a remote patient rhythm surveillance pathway after catheter ablation. The high positive predictive value of ‘Sinus Rhythm’ and ‘AF’ labels and the high specificity of the ‘AF’ label are comparable to diagnostic algorithms in implantable loop recorders.(2) The source code to automate data curation and structured databasing of wearable ECGs has been made available for care providers to integrate the service into device monitoring workflows.(3)

Machine learning to predict incident atrial fibrillation in patients with acute stroke: FIND-AF STROKE

Abstract Background Atrial fibrillation (AF)-related strokes have a high rate of recurrence, and are associated with substantial morbidity, disability, healthcare expenditure, and mortality.(1-3) Detection of AF promises opportunity to prevent recurrent stroke through conversion of antiplatelet to anticoagulant therapy,(4,5) but use of extended monitoring has cost implications.(6) Thus accurately identifying patients with stroke at high risk for AF is important but previous previous models have been developed in small, unrepresentative cohorts.(7) Purpose To develop and validate a simple novel decision support tool for prediction of incident AF after hospitalisation with stroke through machine learning. Methods We developed the model (FIND-AF STROKE) in UK secondary care and death certificate-linked primary care health record data from individuals aged ≥30 years without known AF hospitalized with a stroke or transient ischaemic attack (Jan 2, 1998 to Feb 28, 2022), randomly divided into training (80%) and testing (20%) datasets. We evaluated logistic regression and random forest models for prediction of 2-year AF risk in the testing dataset with internal bootstrap validation, and compared against the CHA2DS2VASC score. Youden index was used to set a threshold for higher and lower risk. Results Of 35 530 UK individuals (mean age 72.1 (SD 12.7) years, 50.9% women), 1742 (4.8%) had developed incident AF at 2 years. Using feature importance a parsimonious version of the model was created requiring only age, sex, ethnicity (white versus other) and five comorbidities. In the testing dataset prediction performance for the random forest algorithm (area under the receiver operating characteristic (AUROC) 0.710, 95% CI 0.685 – 0.736) was superior to the CHA2DS2VASC score (0.645, 95% CI 0.617-0.672) (Figure 1). The threshold for higher risk using the Youden index was 4.81% which resulted in a negative predictive value of 97.6%, sensitivity 0.712, specificity 0.583. Of the post-stroke patients 44% were labelled at higher risk of AF, who were older, had a higher prevalence of comorbidities, and a 3.3-fold higher incidence of AF post-stroke, than those identified as lower risk (Table 1). Conclusions The machine learning FIND-AF STROKE decision support tool can accurately identify individuals at risk of incident AF after stroke to enable a targeted and cost-effective approach to extended monitoring. Validation within cohorts that have undergone extended monitoring post-stroke will give a more accurate estimation of yield at different risk thresholds.Figure 1

Performance of apple watch for detecting atrial fibrillation among a known atrial fibrillation population

Abstract Background Photoplethysmography (PPG) based consumer accessible devices with algorithms for irregular rhythm detection are gaining popularity as an opportunistic screening tool for atrial arrhythmias, and are FDA cleared for use in patients without a history of atrial fibrillation (AF). However, these devices are increasingly used for the management of AF, a purpose for which some algorithms are not intended. Purpose To compare the performance of the Apple Watch Irregular Rhythm Notification (IRN) with an implantable cardiac monitor (ICM) for the detection of atrial fibrillation (AF) in patients with known AF from the DEFINE AFib study. Methods Patients with known AF, a LINQ or LINQ II ICM, and an Apple Watch with PPG-based IRN activated were included in the analysis. ICM detected AF episodes served as the standard to which Apple Watch IRNs were compared. All AF episodes detected by the ICM device were manually adjudicated by two trained, independent reviewers with a third independent reviewer in situations of disagreement. Only AF episodes ≥60 minutes during a 75-minute window in which the Apple Watch was monitoring were included in accordance with the known Apple Watch algorithm detection parameters. To eliminate oversampling of ICM detected episodes compared to Apple Watch IRNs, 1,000 simulations of one randomly selected episode per subject was done to avoid single-subject bias. Results In total, 97 subjects were analyzed of which 42 had an AF episode ≥60 minutes. Applying the 75-minute window of Apple Watch monitoring to episodes ≥60 minutes, 20 subjects had qualifying, adjudicated ICM detected AF. These 20 subjects accounted for 481 qualifying AF episodes. Of the ICM detected episodes, 191 (40%) occurred while an Apple Watch was not being worn. In total, 11/20 (55%) subjects had an IRN generated by the Apple Watch. Adjudication of the ICM episodes resulted in 290 true positive AF episodes ≥60 minutes during a 75-minute window while the watch was actively monitoring, with 74 (26%) of those episodes triggering an IRN resulting in a per episode sensitivity of 24 ± 7% and PPV of 91 ± 4% (Figure 1A). The efficacy of Apple Watch to detect AF episodes of differing lengths was assessed for episodes between 1-6 hours, 6-12 hours, 12-24 hours, and &amp;gt; 24 hours with sensitivities of 19 ± 19%, 38 ± 8%, 32 ± 10%, and 31 ± 8%, respectively (Figure 1B). The Apple Watch generated an annualized false positive rate of 2.25 IRNs for every patient year of monitoring. At the subject level, Apple Watch had a sensitivity of 48%, specificity of 100%, PPV of 100%, and NPV of 71% (Figure 1A). Conclusions In patients with a history of AF, the Apple Watch IRN was most sensitive at the individual subject level while watch performance was lower by episode. The current IRN algorithm favors longer episode durations. The limited sensitivity, dependence on user wear-time, and reliance on long-durations of AF limit the use of Apple Watch for AF management.

Smartphone motion sensor based detection of atrial fibrillation in populations referred to cardiology

Abstract Introduction The motion sensors of a common smartphone when placed on chest can be used to detect vibrations caused by cardiac motion. The concept has been shown to detect atrial fibrillation (AFib) in clinical setting with 96% accuracy. Purpose We sought to examine how the smartphone motion sensor based AFib detection performs in various populations referred to cardiology pooled from 3 different studies addressing arrhythmias, heart failure or suspected coronary artery disease. Methods From a total of 1 524 patient recordings 748 recordings were from men (49%). The mean age of the patients was 64 years (range 20-94 years) and mean BMI was 28.9kg/m2 (range 15.2-63.7kg/m2). The signals were collected by placing a smartphone on chest of the patients in supine position and then analyzed with the CardioSignal solution. From the recordings two 60 second strips were analyzed. A simultaneous 5-lead or 1-lead ECG data was collected for reference. The ECG signals were analyzed for quality and rhythm diagnostics by 2 physicians blinded for the motion sensor AFib detection result. Finally, a head-to-head comparison was conducted for the motion sensor AFib detection result and the ECG data. Results From the 1524 patient recordings the final analysis set included 237 individuals with AFib, whereas the rest had sinus rhythm. The sensitivity to detect AFib was 91.9% and specificity 98.7%. The positive predictive value was 91.0% (88.3% to 94.2%) and the negative predictive value 98.3% (95.2% to 99.1%). The accuracy was 97.0% (94.7% to 97.7%). Conclusions A smartphone-based assessment of AFib using the embedded motion sensors seem to yield very good detection results, especially specificity. The accuracy of the results is comparable to available ECG and PPG based AFib detection solutions.