Atrial Fibrillation Algorithm Research Articles

Detection of atrial fibrillation in persons aged 65years and above using amobile electrocardiogram device.

Untreated atrial fibrillation (AF) often results in increased morbidity and mortality. Opportunistic AF screening in persons aged ≥ 65years is recommended to identify patients with AF in order to prevent AF-related complications. The aim of this study was to assess the feasibility of screening persons for AF with the Kardia mobile electrocardiogram device (MED) and to determine the percentage of newly detected AF cases by selective population screening in the Netherlands. Persons aged ≥ 65years, without amedical history of AF, in nursing homes, at public events or visiting the general practitioner (GP) were approached to participate. AKardia MED smartphone ECG (sECG) was recorded and the CHA2DS2-VASc score was calculated. An automated AF algorithm classified the sECGs as 'sinus rhythm', 'AF' or 'unclassified'. In the case of AF, participants were referred to their GP. All sECGs were assessed by blinded experts. Atotal of 2168 participants were screened for AF. According to the expert's interpretation, 2.5% had newly detected AF, of whom 76.4% never experienced palpitations and 89.1% had aCHA2DS2-VASc score ≥ 2. The algorithm result was unclassified in 12.2% of cases, of which 95.5% were interpretable by experts. With expert opinion as the gold standard and excluding unclassified sECGs, the Kardia MED's negative and positive predictive value for detecting AF was 99.8% and 60.0%, respectively. Screening for AF using the Kardia MED is feasible and results in 2.5% newly detected AF cases. Expert interpretation of algorithm outcomes AF and unclassified is recommended.

Abstract 18672: Impact of Enhanced Artificial Intelligence on Clinic Burden From False Alerts of Insertable Cardiac Monitors

Background: AccuRhythm AI is a suite of cloud-based classification algorithms that filter and suppress false pause alerts and atrial fibrillation (AF) episodes transmitted by insertable cardiac monitors (ICMs). An enhancement to the AF algorithm (AI v2) reduces false AF alerts by 50% relative to the original AI (AI v1) but the impact of this improvement on clinic experience is unknown. Aim: Determine the projected annual impact of AI v2 on Pause and AF alerts in a real-world patient population. Methods: Real-world data were obtained from de-identified LINQ II ICM patients with ≥3 months of follow-up. Previously published performance of each AI algorithm broken out by patient reason for monitor were projected onto the data. Results were extrapolated to the average 200-ICM clinic. The study endpoints were true alert retention, false alert suppression, and clinic review burden reduction. Results: The dataset consisted of 16,301 ICMs with mean follow-up of 7 months. Reasons for monitor were Known AF (25%), Suspected AF / Cryptogenic Stroke (40%), Syncope (23%), and Other (12%). The mean number of alerts per patient-year was 17.9 AF and 4.3 Pause. Projection of AI v2 resulted in minimal loss (0.9%) of true AF and Pause alerts, annually, with no loss in Pause and 1.0% in AF. AI v2 reduced annualized false AF and Pause alerts by 91% (97.4% false Pause and 88.1% false AF reduction) relative to the ICM and 35% relative to AI v1. This translates to 401 and 22 annual review hours saved for a 200-ICM clinic, respectively. Conclusion: The AF enhancement provides a meaningful reduction to the average ICM clinic’s review burden by eliminating most false Pause and AF alerts while maintaining nearly all true alerts.

Blood pressure in atrial fibrillation and in sinus rhythm during ambulatory blood pressure monitoring: data from the TEMPLAR project

The coexistence of hypertension and atrial fibrillation (AF) is common and accounts for a worse prognosis. Uncertainties exist regarding blood pressure (BP) measurements in AF patients by automated oscillometric devices. The Microlife WatchBP 03 AFIB ambulatory BP monitoring (ABPM) device including an AF algorithm with each measurement was used in 430 subjects aged >65 years referred for ABPM and with assumed paroxysmal AF to perform intra-individual comparisons of BP during both AF-indicated and sinus rhythm. Only subjects with >30% of measurements indicating AF and episodes >30 min for assumed AF and for sinus rhythm were included. Mean age was 78 ± 7 years, 43% were male, 77% hypertensive, and 72% were treated. Compared to sinus rhythm, 24-h mean arterial pressure was similar (87.2 ± 9.5 vs 87.5 ± 10.6 mm Hg, p = 0.47), whereas 24-h systolic BP tended to be lower (123.6 ± 13.9 vs 124.7 ± 16.1 mm Hg, p = 0.05) and night-time diastolic BP higher (64.6 ± 10.9 vs 63.3 ± 10.4 mm Hg, p = 0.01) in assumed AF. Diastolic (not systolic) BP variability was higher in AF (p < 0.001). Results were similar with heart rates <90 and ≥90 bpm. In conclusion, this is the first study to use intra-individual comparisons of averaged BP during an ABPM in assumed paroxysmal AF and sinus rhythm. Our results imply that ABPM is feasible and informative also in patients with AF. We also suggest that an AF detection algorithm offers a new approach to evaluate the reliability of averaged BP values in AF compared to SR during an ABPM.

Evaluating atrial fibrillation artificial intelligence for the ED: statistical and clinical implications

ObjectiveAn artificial intelligence (AI) algorithm has been developed to detect the electrocardiographic signature of atrial fibrillation (AF) present on an electrocardiogram (ECG) obtained during normal sinus rhythm. We evaluated the ability of this algorithm to predict incident AF in an emergency department (ED) cohort of patients presenting with palpitations without concurrent AF. MethodsThis retrospective study included patients 18 years and older who presented with palpitations to one of 15 ED sites and had a 12‑lead ECG performed. Patients with prior AF or newly diagnosed AF during the ED visit were excluded. Of the remaining patients, those with a follow up ECG or Holter monitor in the subsequent year were included. We evaluated the performance of the AI-ECG output to predict incident AF within one year of the index ECG by estimating an area under the receiver operating characteristics curve (AUC). Sensitivity, specificity, and positive and negative predictive values were determined at the optimum threshold (maximizing sensitivity and specificity), and thresholds by output decile for the sample. ResultsA total of 1403 patients were included. Forty-three (3.1%) patients were diagnosed with new AF during the following year. The AI-ECG algorithm predicted AF with an AUC of 0.74 (95% CI 0.68–0.80), and an optimum threshold with sensitivity 79.1% (95% Confidence Interval (CI) 66.9%–91.2%), and specificity 66.1% (95% CI 63.6%–68.6%). ConclusionsWe found this AI-ECG AF algorithm to maintain statistical significance in predicting incident AF, with clinical utility for screening purposes limited in this ED population with a low incidence of AF.

Intelligent Algorithm-Based Electrocardiography to Predict Atrial Fibrillation after Coronary Artery Bypass Grafting in the Elderly.

The objective of this study was to explore the predictive value of electrocardiogram (ECG) based on intelligent analysis algorithm for atrial fibrillation (AF) in elderly patients undergoing coronary artery bypass grafting (CABG). Specifically, 106 elderly patients with coronary heart disease who underwent CABG in the hospital were selected, including 52 patients with postoperative AF (AF group) and 54 patients without arrhythmia (control group). Within 1-3 weeks after operation, the dynamic ECG monitoring system based on Gentle AdaBoost algorithm constructed in this study was adopted. After the measurement of the 12-lead P wave duration, the maximum P wave duration (Pmax) and minimum P wave duration (Pmin) were recorded. As for simulation experiments, the same data was used as the back-propagation algorithm. The results showed that for the detection accuracy of the test samples, the Gentle AdaBoost algorithm showed 93.7% accuracy after the first iteration, and the Gentle AdaBoost algorithm was 16.1% higher than the back-propagation algorithm. Compared with the control group, the detection rate of arrhythmia in patients after CABG was significantly lower (P < 0.05). Bivariate logistic regression analysis on Pmax and Pmin showed as follows: Pmax: 95% confidential interval (CI): 1.024-1.081, P < 0.05; Pmin: 95% CI: 1.036-1.117, P < 0.05. The sensitivity of Pmax and Pmin in predicting paroxysmal AF was 78.2% and 73.4%, respectively; the specificity of them was 80.1% and 85.6%, respectively; the positive predictive value was 81.2% and 83.4%, respectively; and the negative predictive value was 79.5% and 75.3%, respectively. In conclusion, the generalization ability of Gentle AdaBoost algorithm was better than that of back-propagation algorithm, and it can identify arrhythmia better. Pmax and Pmin were important indicators of AF after CABG.

Pregled smernica evropskog udruženja kardiologa 2020. za dijagnostikovanje i lečenje atrijalne fibrilacije

Atrial fibrillation (AF) is a complex disease associated with an increased risk of stroke, heart failure, repeated hospitalizations and reduced quality of life, as well as increased mortality. The treatment of AF is complex and requires the involvement of both, the doctor and the entire health system, as well as the patient himself. All international associations of cardiologists periodically publish guidelines for the diagnosis and treatment of AF that are in line with the latest scientific evidence. The latest guidelines of the European Society of Cardiology (ESC) for the treatment and diagnosis of AF, published in 2020, have formulated a diagram of diagnosis and treatment of AF with the aim of facilitating timely diagnosis of AF and optimal treatment of patients with this arrhythmia. The diagram includes diagnostic confirmation that it is AF, structured characterization of patients with AF (4S-AF scheme) and ABC algorithm that includes anticoagulant therapy to prevent thromboembolic complications of AF, better control of symptoms (control of heart rate and / or ventricular rate) as well as detection and treatment of comorbidities and cardiovascular risk factors and correction of harmful life habits. To make it easier to remember the key components of the diagnosis and treatment of AF, the diagram is summarized in the abbreviation CC To ABC, according to English names. The aim of this paper is to present the key messages of the latest ESC guidelines for AF treatment and diagnosis.

Abstract 9757: Frequency of Atrial Fibrillation in a Contemporary Elderly Primary Care Population: A Vital-af Trial Study

Introduction: Reliable estimates of atrial fibrillation (AF) frequency can inform our understanding of population disease burden and shape interventions to reduce AF-related morbidity. Methods: We quantified the prevalence and incidence of AF in VITAL-AF (NCT035115057), a cluster-randomized trial of AF screening using single-lead handheld electrocardiograms within 16 primary care practices affiliated with Massachusetts General Hospital. Potential AF events were identified using the electronic health record. AF events satisfying a high specificity electronic algorithm for prevalent AF (positive predictive value 98.4%) were classified as prevalent AF, while all other potential AF events (including all incident AF events) were manually adjudicated. We calculated AF prevalence and incidence overall and within clinical strata, among individuals who were not screened. Results: Of 18,812 individuals enrolled but not screened between 7/31/2018-10/8/2019 (mean age 74.6±7.2 years, 56.3% women), 2,503 (13.3%) had prevalent AF. AF prevalence was particularly high among individuals aged ≥85 years (29.1%) and individuals with heart failure (43.4%). The AF incidence rate among 16,309 individuals without prevalent AF was 20.1 per 1,000 person-years (95% CI 20.9-26.4). AF incidence also increased markedly with age (e.g., 13.6 at ages 65-69 to 37.1 at ages ≥85 years, p&lt;0.01), was greater in men versus women (26.9 vs. 15.0, p&lt;0.01), and was particularly high among individuals with heart failure (30.7, 95% CI 43.8-69.0). Detailed estimates of AF prevalence and incidence across clinical strata are shown in the Table . Conclusions: Using rigorous methodology within a contemporary, elderly primary care population, we observed the highest reported prevalence of diagnosed AF, including nearly a third of individuals aged 85 years and older. The incidence of new AF is also substantial. Preventive efforts are warranted to reduce the risk of AF and associated complications.

Abstract 11369: Leveraging an Artificial Intelligence-Enabled Ecg Algorithm for Atrial Fibrillation Detection to Predict Long-Term Ischemic Stroke and Cardiovascular Mortality in the Community

Background: Artificial intelligence (AI)-based analysis of the electrocardiogram (ECG) can estimate the probability of having atrial fibrillation (AF) while in sinus rhythm. We evaluated the association between the probability of AF using AI-ECG and incident ischemic stroke (I-Stroke) and cardiovascular (CV) mortality. Methods: We evaluated a community-based cohort of consecutive patients that sought primary care in Olmsted County, MN, from the years 1998 to 2000 followed-up through March 2016. Inclusion criteria were ages 40-79, an electrocardiogram within 2 years and complete CHA2DS2-VASc data. We excluded those with known atherosclerotic CV disease or AF. The AI network was developed in 180,000 patients (649,931 ECG). The primary endpoint was incident I-Stroke; secondary endpoint was I-Stroke or CV mortality; validated in duplicate. Multivariate cox models adjusted for age, sex, history of hypertension, diabetes, heart failure, and in CHA2DS2-VASc Score subgroups (categorized as low=0, moderate=1, and high risk ≥2). Results: We included 14,601 patients, mean age 53.2±13.08 (51% women, 95% whites), CHA2DS2-VASc score 1.3±1.2, median follow-up of 16.3 years, IQR 14.5-17.5, 727 (4.9%) had an I-Stroke and 1309 (8.9%) had I-Stroke or died from CV causes. When comparing the 5 th vs the 1 st quintile, those with high AI-ECG AF probability had increased risk for I-Stroke (HR 3.25, 95%CI 2.55-4.14, P-for trend &lt;.0001 Figure-A ) and ischemic stroke or CV mortality (HR 5.73, 95%CI 4.69-6.99, P-for trend &lt;.0001 Figure-B ). This remained significant after adjustments. Furthermore, AI-ECG AF probability was associated with I-Stroke across CHA2DS2-VASc risk levels (HR 2.43 for high, 2.33 for moderate, 1.98 for low CHA2DS2-VASc risk, all P&lt;0.01). Conclusion: The AI-ECG AF algorithm predicts long-term I-Stroke and CV mortality and may refine CHA2DS2-VASc. These results highlight the utility of the AI enabled ECG when estimating I-Stroke or CV mortality risk.

Interpretable machine learning for early neurological deterioration prediction in atrial fibrillation-related stroke

We aimed to develop a novel prediction model for early neurological deterioration (END) based on an interpretable machine learning (ML) algorithm for atrial fibrillation (AF)-related stroke and to evaluate the prediction accuracy and feature importance of ML models. Data from multicenter prospective stroke registries in South Korea were collected. After stepwise data preprocessing, we utilized logistic regression, support vector machine, extreme gradient boosting, light gradient boosting machine (LightGBM), and multilayer perceptron models. We used the Shapley additive explanation (SHAP) method to evaluate feature importance. Of the 3,213 stroke patients, the 2,363 who had arrived at the hospital within 24 h of symptom onset and had available information regarding END were included. Of these, 318 (13.5%) had END. The LightGBM model showed the highest area under the receiver operating characteristic curve (0.772; 95% confidence interval, 0.715–0.829). The feature importance analysis revealed that fasting glucose level and the National Institute of Health Stroke Scale score were the most influential factors. Among ML algorithms, the LightGBM model was particularly useful for predicting END, as it revealed new and diverse predictors. Additionally, the effects of the features on the predictive power of the model were individualized using the SHAP method.

Interpretable machine learning for early neurological deterioration prediction in atrial fibrillation-related stroke

We aimed to develop a novel prediction model for early neurological deterioration(END) based on an interpretable machine learning(ML) algorithm for atrial fibrillation(AF)-related stroke and to evaluate the prediction accuracy and feature importance of ML models.

A review of hand-held Electrocardiogram (ECG) recording devices

Funding AcknowledgementsType of funding sources: Public hospital(s). Main funding source(s): St Barts NHS TrustIntroductionThe advent of portable hand-held devices to record an electrocardiogram (ECG) in-clinic, has the potential to streamline patient assessment in many clinical environments. These ECG technologies are able to record one to six lead ECGs and enable rapid responses to clinical situations. These devices may also have a role in reducing interdepartmental movement of patients during the COVID-19 pandemic, by facilitating a one-stop shop for clinical evaluation. However despite their benefit, they can be more prone to artefact than a 12-lead ECG.PurposeTo describe the currently available hand-held ECG technologies, their ease of use and diagnostic accuracy for rhythm assessment.MethodAn online review was conducted to identify new ECG technologies. Clinical colleagues were also surveyed for recommendations with respect to specifications, availability of alternative products, and informed of the inclusion /exclusion criteria. The review included devices which had built-in abnormal rhythm notification or display, were cost-effective and easy to use. Devices were automatically excluded if cutaneous skin electrodes were required, no in-built ECG display or notification, additional data transfer required to access ECG data or a high unit cost.ResultsAn initial search uncovered nine devices, with five meeting inclusion /exclusion criteria. These devices were tabulated and compared predefined criteria (See table 1: ECG Technology comparisons). The analysis revealed that all devices utilised two finger placement on their recording electrodes to produce immediate ECG results – no cutaneous ECG application required. The KardiaMobile 6L™ had the largest number of beneficial features, namely multi-channel device, QTc interpretation, rapid diagnosis time and low cost. The intended use of hand-held devices is to detect AF. Four of five devices produced a snapshot ECG for rhythm identification on their display. However, AfibAlert™ devices provided an instant light alert if atrial fibrillation (AF) was detected. Instant check™, Dimitek™ and AfibAlert™ benefit from being complete stand-alone devices. In contrast, KardioMobile devices have a built-in ECG display for instant review however, they require a smart phone or tablet to store and transfer the data. These devices support a reliable internal AF algorithm to obtain a high negative predicted value to safely rule out AF. Frequent premature atrial contractions (PACs) are often difficult to distinguish from AF and can lead to high false-positive rate. Hand-held devices are prone to artefact, however accurate visual assessment able to significantly reduce the amount of ECGs deemed uninterpretable.ConclusionsHand-held ECG technology has potential to become a useful, cost-effective tool during patient consultations, with rapid identification of clinically important arrhythmias. However, limitations exist across providers. A pilot trial of these devices is planned to assess further. Figure.

Detection of Atrial Fibrillation Using a Home Blood Pressure Monitor.

PurposeAtrial fibrillation (AF) is the most common arrhythmia and is associated with an increased risk of complications. A screening test has the potential to prevent AF-related complications. This study investigated the diagnostic accuracy of an automated device for home blood pressure (BP) monitoring, which implements an algorithm for AF detection.Patients and MethodsA modified, automated oscillometric device for home BP monitoring (Omron BP785N (HEM-7321-Z), Omron Healthcare) with an AF detector was used to measure the BP in patients. During each BP measurements, the electrocardiogram (ECG) was recorded simultaneously. Simultaneous BP measurements and ECG recordings were obtained from 99 subjects.ResultsTwenty out of 20 patients with atrial fibrillation were correctly recognized by the device and the device correctly identified 67 patients with sinus rhythm as “Not-AF”. On the other hand, 12 patients with basic rhythm: sinus rhythm were incorrectly referred to as “atrial fibrillation”. In summary, the device has a diagnostic accuracy of 87.88% with a sensitivity of 100% and a specificity of 84.8%. On the other hand, in 23 patients, the raw data of the device showed that a body movement occurred during the measurement of the blood pressure. If these subjects were excluded of the analysis, then the diagnostic accuracy of the device would be even better, namely 90.79%. The sensitivity would be 100% and the specificity 89.5%.ConclusionThese data suggest that an automated device for home blood pressure has an excellent diagnostic accuracy for detecting an AF and could be used as a reliable screening test for early diagnosis of atrial fibrillation. Body movements have an impact of the accuracy and specificity of a blood pressure monitor.

Atrial fibrillation prevalence and risk profile from novel community-based screening in Thailand: A prospective multi-centre study.

BackgroundIn Thailand, almost one-quarter of strokes are related to atrial fibrillation (AF), and many could be prevented if AF were diagnosed and treated prior to the stroke. Therefore, we tested a novel strategy to screen large numbers of community residents using village health volunteers and primary care nurses.MethodsLocal primary care nurses and village health volunteers in Phetchaburi and Lopburi provinces, Thailand were trained to perform AF screening using a blood pressure device with AF algorithm (Microlife A200 AFib). 10% of residents aged ≥ 65 years were randomly selected for screening during home-visits. Participants with possible AF were given follow-up appointments for further testing, including 12-lead ECG and echocardiogram.ResultsOver two-months, 9.7% (13,864/143,478) of the target population were screened: mean age 73.2 ± 6.4 years, 32.4% male. The estimated AF prevalence (detected by Microlife A200 AFib) was 2.8% (95% CI, 2.6–3.1%) for age ≥ 65 years (i.e. 393/13,864 participants). Prevalence increased with age from 1.9% (65–69 years) to 5.0% (≥85 years) (p < 0.001). Only 58% (226/393) of participants with suspected AF attended the follow-up appointment (1–3 months after initial screen): mean CHA2DS2-VASc score 3.2 ± 1.2; 86.3% (195/226) had Class-1 oral anticoagulation recommendation, and 33% (75/226) had AF on 12-lead ECG.ConclusionsIn Thailand, large-scale AF screening in the community is feasible using trained volunteer health workers, allowing screening of large numbers in a short time-period. Further investigation of this strategy is warranted, ensuring mechanisms to obtain a timely rhythm strip or 12-lead ECG locally, and a designated pathway to treatment.

Identification of undiagnosed atrial fibrillation patients using a machine learning risk prediction algorithm and diagnostic testing (PULsE-AI): Study protocol for a randomised controlled trial

Atrial fibrillation (AF) is associated with an increased risk of stroke, enhanced stroke severity, and other comorbidities. However, AF is often asymptomatic, and frequently remains undiagnosed until complications occur. Current screening approaches for AF lack either cost-effectiveness or diagnostic sensitivity; thus, there is interest in tools that could be used for population screening. An AF risk prediction algorithm, developed using machine learning from a UK dataset of 2,994,837 patients, was found to be more effective than existing models at identifying patients at risk of AF. Therefore, the aim of the trial is to assess the effectiveness of this risk prediction algorithm combined with diagnostic testing for the identification of AF in a real-world primary care setting. Eligible participants (aged ≥30 years and without an existing AF diagnosis) registered at participating UK general practices will be randomised into intervention and control arms. Intervention arm participants identified at highest risk of developing AF (algorithm risk score ≥ 7.4%) will be invited for a 12‑lead electrocardiogram (ECG) followed by two-weeks of home-based ECG monitoring with a KardiaMobile device. Control arm participants will be used for comparison and will be managed routinely. The primary outcome is the number of AF diagnoses in the intervention arm compared with the control arm during the research window. If the trial is successful, there is potential for the risk prediction algorithm to be implemented throughout primary care for narrowing the population considered at highest risk for AF who could benefit from more intensive screening for AF.Trial Registration: NCT04045639

Opportunistic screening versus usual care for detection of atrial fibrillation in primary care: cluster randomised controlled trial

ObjectiveTo investigate whether opportunistic screening in primary care increases the detection of atrial fibrillation compared with usual care.DesignCluster randomised controlled trial.Setting47 intention-to-screen and 49 usual care primary care practices in...

The additional value of an algorithm for atrial fibrillation at the stroke unit

Background and purposeThe rate of newly detected (paroxysmal) atrial fibrillation (AF) during inpatient cardiac telemetry is low. The objective of this study was to evaluate the additional diagnostic yield of an automated detection algorithm for AF on telemetric monitoring compared with routine detection by a stroke unit team in patients with recent ischemic stroke or TIA. MethodsPatients admitted to the stroke unit of Medisch Spectrum Twente with acute ischemic stroke or TIA and no history of AF were prospectively included. All patients had telemetry monitoring, routinely assessed by the stroke unit team. The ST segment and arrhythmia monitoring (ST/AR) algorithm was active, with deactivated AF alarms. After 24 h the detections were analyzed and compared with routine evaluation. ResultsFive hundred and seven patients were included (52.5% male, mean age 70.2 ± 12.9 years). Median monitor duration was 24 (interquartile range 22–27) h. In 6 patients (1.2%) routine analysis by the stroke unit team concluded AF. In 24 patients (4.7%), the ST/AR Algorithm suggested AF. Interrater reliability was low (κ, 0.388, p < 0.001). Suggested AF by the algorithm turned out to be false positive in 11 patients. In 13 patients (2.6%) AF was correctly diagnosed by the algorithm. None of the cases detected by routine analysis were missed by the algorithm. ConclusionsAutomated AF detection during 24-h telemetry in ischemic stroke patients is of additional value to detect paroxysmal AF compared with routine analysis by the stroke unit team alone. Automated detections need to be carefully evaluated.

Implementation and validation of real-time algorithms for atrial fibrillation detection on a wearable ECG device

BackgroundDue to the growing epidemic of atrial fibrillation (AF), new strategies for AF screening, diagnosis, and monitoring are required. Wearable devices with on-board AF detection algorithms may improve early diagnosis and therapy outcomes. In this work, we implemented optimized algorithms for AF detection on a wearable ECG monitoring device and assessed their performance. MethodsThe signal processing framework was composed of two main modules: 1) a QRS detector based on a finite state machine, and 2) an AF detector based on the Shannon entropy of the symbolic word series obtained from the instantaneous heart rate. The AF detector was optimized off-line by tuning its parameters to reduce the computational burden while preserving detection accuracy. On-board performance was assessed in terms of detection accuracy, memory usage, and computation time. ResultsThe on-board implementation of the QRS detector produced an overall accuracy of 99% on the MIT-BIH Arrhythmia Database, with memory usage = 672 bytes, and computation time ≤90 μs. The on-board implementation of the optimized AF algorithm gave an overall accuracy of 98.1% (versus 98.3% of the original version) on the MIT-BIH AF Database, with increased sensitivity (99.2% versus 98.5%) and decreased specificity (97.3% versus 98.2%), memory usage = 4648 bytes, and computation time ≤ 75 μs (consistent with real-time detection). ConclusionsThis study demonstrated the feasibility of real-time AF detection on a wearable ECG device. It constitutes a promising step towards the development of novel ECG monitoring systems to tackle the growing AF epidemic.

4118Cisplatin therapy as a risk factor for incident atrial fibrillation

Abstract Background Although aging is a major contributor to the rising epidemic of atrial fibrillation (AF), increased numbers of cancer survivors who suffer from chemotherapy and radiotherapy mediated cardiovascular disease also appear to be especially vulnerable to developing the arrhythmia. Currently, little is known about the long-term risk of AF in patients treated with these two treatment modalities for cancer. Purpose To define the incidence and prevalence of AF in a thoracic tumour cohort following radiotherapy and chemotherapy. Methods We conducted a retrospective study from the electronic health records (EHR) of patients who had a primary tumour of the breast and lung. Patients with AF, radiotherapy or chemotherapy prior to cancer diagnosis were excluded. All patients were followed for incident AF until last censoring or death using a validated AF algorithm (positive predictive value 96%). Radiation oncology treatment reports were reviewed and chemotherapy data was extracted from the EHR and manually reviewed for 43 chemotherapeutic agents. Results In the tumour registry, 17,402 patients were identified by histological and pathological reports. We examined 3,336 patients with a thoracic tumour diagnosis (breast and lung) with a mean age of 62.7±12.5 (SD) years. The AF incident rate for breast and lung cancer was 1.2 and 13.5 cases per every 100 persons with a cumulative rate of 2.49 and 0.67 respectively. We observed an overall AF prevalence of 3.4% and 7.4% in the breast and lung cohorts respectively. The average follow-up duration was 4.69 years (0.0–17.8 years) in breast cancer and 1.36 years (0.0–13.9 years) in lung cancer patients. A Cox regression model revealed no association between radiotherapy and AF in either cancer type (breast: P=0.18 and lung: P=0.16). However, patients who received cisplatin had 4-fold increased risk of developing AF after covariate adjustment (HR = 4.22, 95% CI: 0.10–17.85, P=0.05). Conclusions We show for the first time in a large thoracic tumour registry and using the EHR that patients treated with cisplatin were more likely to develop AF. Additional studies are needed to validate our findings and to further investigate the underlying mechanisms of these cardiotoxic effects. Acknowledgement/Funding This work was supported in part by grants from the AHA 17MCPRP33420153 (Brandon Chalazan) and NIH/NHLBI R01 HL092217 and HL138737 (Dawood Darbar).

Mobile Phone-Based Use of the Photoplethysmography Technique to Detect Atrial Fibrillation in Primary Care: Diagnostic Accuracy Study of the FibriCheck App.

BackgroundMobile phone apps using photoplethysmography (PPG) technology through their built-in camera are becoming an attractive alternative for atrial fibrillation (AF) screening because of their low cost, convenience, and broad accessibility. However, some important questions concerning their diagnostic accuracy remain to be answered.ObjectiveThis study tested the diagnostic accuracy of the FibriCheck AF algorithm for the detection of AF on the basis of mobile phone PPG and single-lead electrocardiography (ECG) signals.MethodsA convenience sample of patients aged 65 years and above, with or without a known history of AF, was recruited from 17 primary care facilities. Patients with an active pacemaker rhythm were excluded. A PPG signal was obtained with the rear camera of an iPhone 5S. Simultaneously, a single‑lead ECG was registered using a dermal patch with a wireless connection to the same mobile phone. PPG and single-lead ECG signals were analyzed using the FibriCheck AF algorithm. At the same time, a 12‑lead ECG was obtained and interpreted offline by independent cardiologists to determine the presence of AF.ResultsA total of 45.7% (102/223) subjects were having AF. PPG signal quality was sufficient for analysis in 93% and single‑lead ECG quality was sufficient in 94% of the participants. After removing insufficient quality measurements, the sensitivity and specificity were 96% (95% CI 89%-99%) and 97% (95% CI 91%-99%) for the PPG signal versus 95% (95% CI 88%-98%) and 97% (95% CI 91%-99%) for the single‑lead ECG, respectively. False-positive results were mainly because of premature ectopic beats. PPG and single‑lead ECG techniques yielded adequate signal quality in 196 subjects and a similar diagnosis in 98.0% (192/196) subjects.ConclusionsThe FibriCheck AF algorithm can accurately detect AF on the basis of mobile phone PPG and single-lead ECG signals in a primary care convenience sample.

Effectiveness and accuracy of arrhythmia detection algorithm with the Reveal LINQ insertable cardiac monitor

Objective: To investigate the accuracy of arrhythmia detection algorithm and the effectiveness of Reveal LINQ insertable cardiac monitor (ICM). Methods: This single-center, descriptive, non-controlled clinical study was designed to enroll consecutive patients who were implanted with Reveal LINQ ICM in Minneapolis Veterans Affairs Medical Center from June 19, 2014 to April 18, 2017. The safety of implant procedure, the accuracy of automatic arrhythmia detection algorithm and the clinical effectiveness were evaluated and validated independently by two physicians. Results: Sixty patients with average age of (65±13) years (30-89 years) were enrolled and the follow-up duration was 466 days (30-1 072 days). Of these, 57 patients are male. The indications of ICM implantation were unexplained syncope in 35 cases, cryptogenic stroke in 13 cases, atrial fibrillation (AF) management in 5 cases, and palpitations in 7 cases. The implantation procedure was safe and successful in all patients without complications. The detection accuracy of arrhythmia algorithm was 86% (620/723), 82% (2 369/2 876), and 30% (411/1 380) for tachycardia, bradycardia, and pause events, respectively. The detection accuracy of arrhythmia algorithm for AF was 17% (5/30), 60% (1 569/2 632) and 69% (172/251) for cryptogenic stroke, unexplained syncope, and AF management, respectively. Important clinical findings were observed in 15% (9/60) patients. Only 5% (3/60) patients had important findings based on the intended clinical indications. Conclusion: The Reveal LINQ ICM is a safe, simple, and effective procedure for the detection of tachycardia and bradycardia events. However, the automatic AF and pause events detection algorithm had high false positive rate which requires further improvement.