Circulation: Arrhythmia and Electrophysiology September 2019 Issue Circulation: Arrhythmia and Electrophysiology On the Beat

Dr Wang:             Welcome to the monthly podcast, On the Beat, for Circulation: Arrhythmia and Electrophysiology. I'm Dr Paul Wang, Editor-in-Chief, with some of the key highlights from this month's issue.
                                In our first paper, Ying Tian and associates examine the effects and long-term outcomes of percutaneous stellate ganglion blockade in the setting of drug refractory electrical storm due to ventricular arrhythmia. They studied 30 consecutive patients over nearly a five-year period. They used bupivacaine alone, or in combination with lidocaine injected into the neck with good local anesthetic spread in the vicinity of the left stellate ganglion in 15 patients, or both stellate ganglion in 15 patients.
                                The mean left ventricular ejection fraction was 34%. At 24 hours, 60% of patients were free of ventricular arrhythmia. Patients whose ventricular arrhythmia was controlled had a lower hospital mortality rate than patients whose ventricular arrhythmia continued. 5.6 versus 50%, P equals 0.009. Implantable cardioverter-defibrillator interrogation showed a significant 92% reduction in ventricular arrhythmia episodes from 26 to 2 in the 72 hours after stellate ganglion blockade, P less than 0.001.
                                Patients who died during the same hospitalization, N equals 7, were more likely to have ischemic cardiomyopathy, 100% versus 43.5%. And recurrent ventricular arrhythmias within 24 hours, 85.7% versus 26.1%. There were no procedure related complications.
                                In our next paper, Zachi Attia and associates hypothesized that a convolutional neural network could be trained through a process called 'deep learning' to predict a person's age and gender using only 12-lead electrocardiogram signals. They trained convolutional neural network using 10 second samples of 12-lead ECG signals from 499,727 patients to predict gender and age. The networks were tested on a separate cohort of 275,056 patients. For gender classification, the model obtained 90.4% classification accuracy with an area under the curve of 0.97. In the independent test data, age was estimated as a continuous variable with an average error of 6.9 years, R squared equals 0.7.
                                Among 100 hundred patients with multiple ECGs over the course of at least two decades of life, most patients, 51%, had an average error between real age and convolutional neural network predicted age of less than seven years. Major factors seen amongst patients with convolutional neural network predicted age that exceeded chronologic age by greater than seven years included low ejection fraction, hypertension, and coronary disease, P less than 0.1. In the 27% of patients whose correlation was greater than 0.8, between convolutional neural network predicted and chronological age, no incident events occurred over follow up 30 years.
                                The authors concluded that applying artificial intelligence to the ECG allows prediction of patient, gender, and estimation of age. The ability of artificial intelligent algorithm to determine physiological age with further validation may serve as a measure of overall health.
                                In our next paper, Zain Ul Abideen Asad and associates performed a meta-analysis of randomized control trials in order to compare the efficacy and safety of catheter ablation with medical therapy for atrial fibrillation with the primary outcome being all-cause mortality. They examined 18 randomized controlled trials comprising 4,464 patients. Catheter ablation resulted in significant reduction in all-cause mortality, relative risk of 0.69 that was driven by