Artificial intelligence augments detection accuracy of cardiac insertable cardiac monitors: Results from a pilot prospective observational study

Fabio Quartieri, Manuel Marina-Breysse, Annalisa Pollastrelli, Isabella Paini, Carlos Lizcano, José María Lillo-Castellano, Andrea Grammatico, Fabio Quartieri, Manuel Marina-Breysse, Annalisa Pollastrelli, Isabella Paini, Carlos Lizcano, José María Lillo-Castellano, Andrea Grammatico

Abstract

Background: Insertable cardiac monitors (ICMs) are indicated for long-term monitoring of patients with unexplained syncope or who are at risk for cardiac arrhythmias. The volume of ICM-transmitted information may result in long data review times to identify true and clinically relevant arrhythmias.

Objective: The purpose of this study was to evaluate whether artificial intelligence (AI) may improve ICM detection accuracy.

Methods: We performed a retrospective analysis of consecutive patients implanted with the Confirm RxTM ICM (Abbott) and followed in a prospective observational study. This device continuously monitors subcutaneous electrocardiograms (SECGs) and transmits to clinicians information about detected arrhythmias and patient-activated symptomatic episodes. All SECGs were classified by expert electrophysiologists and by the WillemTM AI algorithm (IDOVEN).

Results: During mean follow-up of 23 months, of 20 ICM patients (mean age 68 ± 12 years; 50% women), 19 had 2261 SECGs recordings associated with cardiac arrhythmia detections or patient symptoms. True arrhythmias occurred in 11 patients: asystoles in 2, bradycardias in 3, ventricular tachycardias in 4, and atrial tachyarrhythmias (atrial tachycardia/atrial fibrillation [AT/AF]) in 10; with 6 patients having >1 arrhythmia type. AI algorithm overall accuracy for arrhythmia classification was 95.4%, with 97.19% sensitivity, 94.52% specificity, 89.74% positive predictive value, and 98.55% negative predictive value. Application of AI would have reduced the number of false-positive results by 98.0% overall: 94.0% for AT/AF, 87.5% for ventricular tachycardia, 99.5% for bradycardia, and 98.8% for asystole.

Conclusion: Application of AI to ICM-detected episodes is associated with high classification accuracy and may significantly reduce health care staff workload by triaging ICM data.

Keywords: Artificial intelligence; Detection accuracy; Insertable cardiac monitors.

© 2022 Heart Rhythm Society.

Figures

Figure 1
Figure 1
Examples of electrocardiograms from different patients showing events that were correctly classified the WillemTM artificial intelligence algorithm according to the real cardiac alteration (WillemTM TP column). AT/AF = atrial tachycardia/atrial fibrillation (ie, atrial tachyarrhythmia); Brady = bradycardia; TP = true positive; VT = ventricular tachycardia.
Figure 2
Figure 2
Examples of electrocardiograms from different patients showing events that were misclassified (false-positive [FP]) or undetected (false-negative [FN]) by the WillemTM artificial intelligence algorithm (WillemTM FP & FN column). The annotations given by cardiologists also are shown (Cardiologist column). Abbreviations as in Figure 1.
Figure 3
Figure 3
Examples of electrocardiograms from different patients showing events that were misclassified by the WillemTM artificial intelligence algorithm (WillemTM FP column). The annotations of the cardiologists also are shown (Cardiologist column). Abbreviations as in Figure 1, Figure 2.
Figure 4
Figure 4
Histogram showing the absolute number of false-positive episodes detected by the insertable cardiac monitor (ICM) (blue bars) and after application of the WillemTM artificial intelligence algorithm (black bars). The reduction in false-positive detections is also expressed as percentage and indicated next to the arrows for each episode type. Abbreviations as in Figure 1.
Figure 5
Figure 5
Examples of electrocardiograms from different patients showing events that were detected as arrhythmias by the insertable cardiac monitor (ICM FP column) and then correctly reclassified by the WillemTM artificial intelligence algorithm according to the real cardiac alteration made by the cardiology experts considered as the gold standard (WillemTM TP column). Abbreviations as in Figure 1, Figure 2.

References

    1. Shen W.K., Sheldon R.S., Benditt D.G., et al. 2017 ACC/AHA/HRS Guideline for the evaluation and management of patients with syncope: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. Circulation. 2017;136:e60–e122.
    1. Brignole M., Moya A., de Lange F.J., et al. 2018 ESC Guidelines for the diagnosis and management of syncope. Eur Heart J. 2018;39:1883–1948.
    1. Tracy C.M., Epstein A.E., Darbar D., et al. 2012 ACCF/AHA/HRS Focused update of the 2008 guidelines for device-based therapy of cardiac rhythm abnormalities: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Heart Rhythm. 2012;9:1737–1753.
    1. Al-Khatib S.M., Stevenson W.G., Ackerman M.J., et al. 2017 AHA/ACC/HRS Guideline for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. Heart Rhythm. 2018;15:e73–e189.
    1. Ip J., Jaffe B., Castellani M., Sheikh A., Castellani C., Ip R. Accuracy of arrhythmia detection in implantable cardiac monitors: a prospective randomized clinical trial comparing Reveal LINQ and Confirm Rx. Pacing Clin Electrophysiol. 2020;43:1344–1350.
    1. Hindricks G., Pokushalov E., Urban L., et al. Performance of a new leadless implantable cardiac monitor in detecting and quantifying atrial fibrillation results of the XPECT trial. Circ Arrhythm Electrophysiol. 2010;3:141–147.
    1. Pürerfellner H., Pokushalov E., Sarkar S., et al. P-wave evidence as a method for improving algorithm to detect atrial fibrillation in insertable cardiac monitors. Heart Rhythm. 2014;11:1575–1583.
    1. Sanders P., Pürerfellner H., Pokushalov E., et al. Performance of a new atrial fibrillation detection algorithm in a miniaturized insertable cardiac monitor: results from the Reveal LINQ Usability Study. Heart Rhythm. 2016;13:1425–1430.
    1. Mittal S., Rogers J., Sarkar S., et al. Real-world performance of an enhanced atrial fibrillation detection algorithm in an insertable cardiac monitor. Heart Rhythm. 2016;13:1624–1630.
    1. Nolker G., Mayer J., Boldt L.H., et al. Performance of an implantable cardiac monitor to detect atrial fibrillation: results of the DETECT AF Study. J Cardiovasc Electrophysiol. 2016;27:1403–1410.
    1. Pürerfellner H., Sanders P., Sarkar S., et al. Adapting detection sensitivity based on evidence of irregular sinus arrhythmia to improve atrial fibrillation detection in insertable cardiac monitors. Europace. 2018;20:321–328.
    1. Afzal M.R., Mease J., Koppert T., et al. Incidence of false-positive transmissions during remote monitoring with implantable loop recorders. Heart Rhythm. 2020;17:75–80.
    1. Gopinathannair R., Lakkireddy D., Afzal M.R., et al. Effectiveness of SharpSense™ algorithms in reducing bradycardia and pause detection: real-world performance in Confirm Rx™ insertable cardiac monitor. J Interv Card Electrophysiol. 2022;63:661–668.
    1. Hannun A.Y., Rajpurkar P., Haghpanahi M., et al. Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network. Nat Med. 2019;25:65–69.
    1. Smith S.W., Rapin J., Li J., et al. A deep neural network for 12-lead electrocardiogram interpretation outperforms a conventional algorithm, and its physician overread, in the diagnosis of atrial fibrillation. Int J Cardiol Heart Vasc. 2019;25
    1. Attia Z.I., Kapa S., Lopez-Jimenez F., et al. Screening for cardiac contractile dysfunction using an artificial intelligence-enabled electrocardiogram. Nat Med. 2019;25:70–74.
    1. Attia Z.I., Noseworthy P.A., Lopez-Jimenez F., et al. An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction. Lancet. 2019;394:861–867.
    1. Raghunath S., Cerna A.E.U., Jing L., et al. Prediction of mortality from 12-lead electrocardiogram voltage data using a deep neural network. Nat Med. 2020;26:886–891.
    1. Mittal S., Oliveros S., Li J., Barroyer T., Henry C., Gardella C. AI filter improves positive predictive value of atrial fibrillation detection by an implantable loop recorder. JACC Clin Electrophysiol. 2021;7:965–975.
    1. Lillo-Castellano J.M., González-Ferrer J.J., Marina-Breysse M., et al. Personalized monitoring of electrical remodelling during atrial fibrillation progression via remote transmissions from implantable devices. Europace. 2020;22:704–715.
    1. Pouyanfar S., Chen Shu-Ching, ShyuAn Mei-Ling. Efficient deep residual-inception network for multimedia classification. IEEE International Conference on Multimedia and Expo Conference Paper. 2017:373–378. doi: 10.1109/ICME.2017.8019447.
    1. Ferri C., Hernández-Orallo J., Modroiu R. An experimental comparison of performance measures for classification. Pattern Recognition Letters. 2009;30:27–38.

Source: PubMed

Sponsorer og samarbeidspartnere

Medisinsk tilstand

Legemiddelintervensjoner

3
Abonnere