Detection of Non-Sustained Supraventricular Tachycardia in Atrial Fibrillation Screening.

OBJECTIVE: Non-sustained supraventricular tachycardia (nsSVT) is associated with a higher risk of developing atrial fibrillation (AF), and, therefore, detection of nsSVT can improve AF screening efficiency. However, the detection is challenged by the lower signal quality of ECGs recorded using handheld devices and the presence of ectopic beats which may mimic the rhythm characteristics of nsSVT. METHODS: The present study introduces a new nsSVT detector for use in single-lead, 30-s ECGs, based on the assumption that beats in an nsSVT episode exhibits similar morphology, implying that episodes with beats of deviating morphology, either due to ectopic beats or noise/artifacts, are excluded. A support vector machine is used to classify successive 5-beat sequences in a sliding window with respect to similar morphology. Due to the lack of adequate training data, the classifier is trained using simulated ECGs with varying signal-to-noise ratio. In a subsequent step, a set of rhythm criteria is applied to similar beat sequences to ensure that episode duration and heart rate is acceptable. RESULTS: The performance of the proposed detector is evaluated using the StrokeStop II database, resulting in sensitivity, specificity, and positive predictive value of 84.6%, 99.4%, and 18.5%, respectively. CONCLUSION: The results show that a significant reduction in expert review burden (factor of 6) can be achieved using the proposed detector.Clinical and Translational Impact: The reduction in the expert review burden shows that nsSVT detection in AF screening can be made considerably more efficiently.

An artificial intelligence-based model for prediction of atrial fibrillation from single-lead sinus rhythm electrocardiograms facilitating screening.

AIMS: Screening for atrial fibrillation (AF) is recommended in the European Society of Cardiology guidelines. Yields of detection can be low due to the paroxysmal nature of the disease. Prolonged heart rhythm monitoring might be needed to increase yield but can be cumbersome and expensive. The aim of this study was to observe the accuracy of an artificial intelligence (AI)-based network to predict paroxysmal AF from a normal sinus rhythm single-lead ECG. METHODS AND RESULTS: A convolutional neural network model was trained and evaluated using data from three AF screening studies. A total of 478 963 single-lead ECGs from 14 831 patients aged ≥65 years were included in the analysis. The training set included ECGs from 80% of participants in SAFER and STROKESTOP II. The remaining ECGs from 20% of participants in SAFER and STROKESTOP II together with all participants in STROKESTOP I were included in the test set. The accuracy was estimated using the area under the receiver operating characteristic curve (AUC). From a single timepoint ECG, the artificial intelligence-based algorithm predicted paroxysmal AF in the SAFER study with an AUC of 0.80 [confidence interval (CI) 0.78-0.83], which had a wide age range of 65-90+ years. Performance was lower in the age-homogenous groups in STROKESTOP I and STROKESTOP II (age range: 75-76 years), with AUCs of 0.62 (CI 0.61-0.64) and 0.62 (CI 0.58-0.65), respectively. CONCLUSION: An artificial intelligence-enabled network has the ability to predict AF from a sinus rhythm single-lead ECG. Performance improves with a wider age distribution.

Factors predicting participation and potential yield of screening-detected disease among non-participants in a Swedish population-based atrial fibrillation screening study.

BACKGROUND: The success of any screening program is dependent on participation. The characteristics of participants vs. non-participants have been studied and non-participants usually have a higher risk of disease. The potential yield of screening-detected disease in non-participants could be of interest to several screening programs. AIMS: This is a sub-study to STROKESTOP II, a Swedish atrial fibrillation screening study. The aim was to study factors predicting participation and to estimate the potential yield of screening-detected disease in non-participants. METHODS: Individual, anonymized data for participants and non-participants with respect to socioeconomic factors, medical history and drugs dispensed were obtained from Swedish registries. A random forest model was trained to predict propensity scores for participation. The propensity scores were used to estimate potential screening-detected disease among non-participants. RESULTS: Non-participants (n = 7086) had lower income, were more likely to have been hospitalized and had higher CHA2DS2-VASc scores compared to participants (n = 6868). The strongest factor predicting non-attendance was low income. The weighted estimates suggested that the yield of new atrial fibrillation was 2.4% in non-participants compared to 2.3% in the participants, which was not significant. CONCLUSIONS: Non-participants had higher CHA2DS2-VASc scores, indicating a higher stroke-risk and presumable benefit from attending screening, although estimated new atrial fibrillation detected was not significantly more common when compared to participants. Low income was the strongest factor for predicting non-attendance and should be a focus area when planning future screening scenarios.

Prognostic Implications of Supraventricular Arrhythmias.

The aim of this study was to establish the prevalence and prognostic implication of progressive supraventricular arrhythmias from frequent supraventricular ectopic complexes, isolated, in bi- or trigeminy, to supraventricular tachycardias with different characteristics. In the STROKESTOP I mass-screening study for atrial fibrillation (AF) in 75- and 76-year olds in Sweden, participants registered 30-second intermittent ECG twice daily for two weeks. The ECG-recordings from STROKESTOP I were re-evaluated using an automated algorithm to detect individuals with frequent supraventricular ectopic complexes or runs. Detected episodes were manually re-examined to confirm the findings. The primary endpoint was AF as ascertained from the national Swedish Patient register. Exploratory secondary endpoints were stroke and death. Median follow-up was 4.2 (interquartile range [IQR] 3.8-4.4) years. Of the examined 6,100 participants, 85% were free of significant supraventricular arrhythmia. In the 894 participants that had arrhythmia, frequent supraventricular ectopic complexes were the most common arrhythmia, n = 709 (11.6%) and irregular supraventricular tachycardias were more common than regular. Individuals with the most AF similar supraventricular tachycardias, irregular and lacking p-waves (termed micro-AF), n = 97 (1.6%) had the highest risk of developing AF (hazard ratio 4.3; 95% confidence interval [CI] 2.7-6.8). They also had increased risk of death (hazard ratio 2.0; CI 1.1-3.8). In conclusion, progression of atrial arrhythmias from supraventricular ectopic complexes to more AF-like episodes is associated with development of AF. Extended screening for AF should be considered in individuals with frequent supraventricular activity, especially in those with supraventricular tachycardias with AF characteristics.