Erkennung, Vorhersage und Behandlung von Vorhofflimmern mithilfe künstlicher Intelligenz

Jonas L. Isaksen1, Mathias Baumert2, Astrid N L Hermans3, Mary M. Maleckar4, Dominik Linz1,3
1Department of Biomedical Sciences, University of Copenhagen, Copenhagen, Denmark
2School of Electrical and Electronic Engineering, The University of Adelaide, Adelaide, Australia
3Department of Cardiology, Maastricht University Medical Center and Cardiovascular Research Institute Maastricht, Maastricht, The Netherlands
4Department of Computational Physiology, Simula Research Laboratory, Oslo, Norway

Tóm tắt

Abstract

The present article reviews the state of the art of machine learning algorithms for the detection, prediction, and management of atrial fibrillation (AF), as well as of the development and evaluation of artificial intelligence (AI) in cardiology and beyond. Today, AI detects AF with a high accuracy using 12-lead or single-lead electrocardiograms or photoplethysmography. The prediction of paroxysmal or future AF currently operates at a level of precision that is too low for clinical use. Further studies are needed to determine whether patient selection for interventions may be possible with machine learning.

Từ khóa


Tài liệu tham khảo

Hindricks G, Potpara T, Dagres N, Arbelo E, Bax JJ, Blomström-Lundqvist C, Boriani G, Castella M, Dan G‑A, Dilaveris PE et al (2021) 2020 ESC guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European association for cardio-thoracic surgery (EACTS): the task force for the diagnosis and management of atrial fibrillation of the European society of cardiology (ESC) developed with the special contribution of the European heart rhythm association (EHRA) of the ESC. Eur Heart J 42:373–498

Reinhold T, Lindig C, Willich SN, Brüggenjürgen B (2011) The costs of atrial fibrillation in patients with cardiovascular comorbidities—a longitudinal analysis of German health insurance data. Europace 13:1275–1280

Schnabel RB, Wallenhorst C, Engler D, Blankenberg S, Pfeiffer N, Spruenker NA, Buettner M, Michal M, Lackner KJ, Münzel T et al (2021) Refined atrial fibrillation screening and cost-effectiveness in the German population. Heart. https://doi.org/10.1136/heartjnl-2020-318882

Hannun AY, Rajpurkar P, Haghpanahi M, Tison GH, Bourn C, Turakhia MP, Ng AY (2019) Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network. Nat Med 25:65–69

Nagendran M, Chen Y, Lovejoy CA, Gordon AC, Komorowski M, Harvey H, Topol EJ, Ioannidis JPA, Collins GS, Maruthappu M (2020) Artificial intelligence versus clinicians: systematic review of design, reporting standards, and claims of deep learning studies. BMJ 368:m689

Hicks SA, Isaksen JL, Thambawita V, Ghouse J, Ahlberg G, Linneberg A, Grarup N, Strümke I, Ellervik C, Olesen MS et al (2021) Explaining deep neural networks for knowledge discovery in electrocardiogram analysis. Sci Rep 11:10949

Zech JR, Badgeley MA, Liu M, Costa AB, Titano JJ, Oermann EK (2018) Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: a cross-sectional study. PLoS Med 15:e1002683

Lin H, Li R, Liu Z, Chen J, Yang Y, Chen H, Lin Z, Lai W, Long E, Wu X et al (2019) Diagnostic efficacy and therapeutic decision-making capacity of an artificial intelligence platform for childhood cataracts in eye clinics: a multicentre randomized controlled trial. EClinicalMedicine 9:52–59

Hill NR, Sandler B, Mokgokong R, Lister S, Ward T, Boyce R, Farooqui U, Gordon J (2020) Cost-effectiveness of targeted screening for the identification of patients with atrial fibrillation: evaluation of a machine learning risk prediction algorithm. J Med Econ 23:386–393

McCarthy J, Minsky ML, Rochester N, Corporation IBM, Shannon CE A proposal for the Dartmouth summer research project on artificial intelligence. http://jmc.stanford.edu/articles/dartmouth.html. Accessed 13 Jan 2022

Turing AM (1950) Computing machinery and intelligence. Mind LIX:433–460

LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:436–444

Attia ZI, Noseworthy PA, Lopez-Jimenez F, Asirvatham SJ, Deshmukh AJ, Gersh BJ, Carter RE, Yao X, Rabinstein AA, Erickson BJ et al (2019) An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction. Lancet 394:861–867

Ozenne B, Subtil F, Maucort-Boulch D (2015) The precision–recall curve overcame the optimism of the receiver operating characteristic curve in rare diseases. J Clin Epidemiol 68:855–859

Altman DG, Bland JM (1983) Measurement in medicine: the analysis of method comparison studies. J R Stat Soc Ser A Stat Soc 32:307–317

Cai W, Chen Y, Guo J, Han B, Shi Y, Ji L, Wang J, Zhang G, Luo J (2020) Accurate detection of atrial fibrillation from 12-lead ECG using deep neural network. Comput Biol Med 116:103378

Jo Y‑Y, Cho Y, Lee SY, Kwon J‑M, Kim K‑H, Jeon K‑H, Cho S, Park J, Oh B‑H (2021) Explainable artificial intelligence to detect atrial fibrillation using electrocardiogram. Int J Cardiol 328:104–110

Baalman SWE, Schroevers FE, Oakley AJ, Brouwer TF, van der Stuijt W, Bleijendaal H, Ramos LA, Lopes RR, Marquering HA, Knops RE et al (2020) A morphology based deep learning model for atrial fibrillation detection using single cycle electrocardiographic samples. Int J Cardiol 316:130–136

Mittal S, Oliveros S, Li J, Barroyer T, Henry C, Gardella C (2021) AI filter improves positive predictive value of atrial fibrillation detection by an implantable loop recorder. JACC Clin Electrophysiol 7:965–975

Hermans ANL, Gawalko M, Dohmen L, van der Velden RMJ, Betz K, Duncker D, Verhaert DVM, Heidbuchel H, Svennberg E, Neubeck L et al (2021) Mobile health solutions for atrial fibrillation detection and management: a systematic review. Clin Res Cardiol. https://doi.org/10.1007/s00392-021-01941-9

Clifford GD, Liu C, Moody B, Lehman L‑WH, Silva I, Li Q, Johnson AE, Mark RG (2017) AF classification from a short single lead ECG recording: the physionet/computing in cardiology challenge 2017. Comput Cardiol. https://doi.org/10.22489/CinC.2017.065-469

Rad AB, Galloway C, Treiman D, Xue J, Li Q, Sameni R, Albert D, Clifford GD (2021) Atrial fibrillation detection in outpatient electrocardiogram monitoring: an algorithmic crowdsourcing approach. PLoS ONE 16:e259916

Zhang P, Ma C, Sun Y, Fan G, Song F, Feng Y, Zhang G (2021) Global hybrid multi-scale convolutional network for accurate and robust detection of atrial fibrillation using single-lead ECG recordings. Comput Biol Med 139:104880

Zhang X, Li J, Cai Z, Zhang L, Chen Z, Liu C (2021) Over-fitting suppression training strategies for deep learning-based atrial fibrillation detection. Med Biol Eng Comput 59:165–173

Oster J, Hopewell JC, Ziberna K, Wijesurendra R, Camm CF, Casadei B, Tarassenko L (2020) Identification of patients with atrial fibrillation: a big data exploratory analysis of the UK biobank. Physiol Meas 41:25001

Wasserlauf J, You C, Patel R, Valys A, Albert D, Passman R (2019) Smartwatch performance for the detection and quantification of atrial fibrillation. Circ Arrhythm Electrophysiol 12:e6834

Tutuko B, Nurmaini S, Tondas AE, Rachmatullah MN, Darmawahyuni A, Esafri R, Firdaus F, Sapitri AI (2021) AFibNet: an implementation of atrial fibrillation detection with convolutional neural network. BMC Med Inform Decis Mak 21:216

Teplitzky BA, McRoberts M, Ghanbari H (2020) Deep learning for comprehensive ECG annotation. Heart Rhythm 17:881–888

Betz K, van der Velden R, Gawalko M, Hermans A, Pluymaekers N, Hillmann HAK, Hendriks J, Duncker D, Linz D (2021) Interpretation der Photoplethysmographie: Schritt für Schritt. Herzschrittmacherther Elektrophysiol 32:406–411

Chen E, Jiang J, Su R, Gao M, Zhu S, Zhou J, Huo Y (2020) A new smart wristband equipped with an artificial intelligence algorithm to detect atrial fibrillation. Heart Rhythm 17:847–853

Tison GH, Sanchez JM, Ballinger B, Singh A, Olgin JE, Pletcher MJ, Vittinghoff E, Lee ES, Fan SM, Gladstone RA et al (2018) Passive detection of atrial fibrillation using a commercially available smartwatch. JAMA Cardiol 3:409–416

Aschbacher K, Yilmaz D, Kerem Y, Crawford S, Benaron D, Liu J, Eaton M, Tison GH, Olgin JE, Li Y et al (2020) Atrial fibrillation detection from raw photoplethysmography waveforms: a deep learning application. Heart Rhythm O2 1:3–9

Poh M‑Z, Poh YC, Chan P‑H, Wong C‑K, Pun L, Leung WW‑C, Wong Y‑F, Wong MM‑Y, Chu DW‑S, Siu C‑W (2018) Diagnostic assessment of a deep learning system for detecting atrial fibrillation in pulse waveforms. Heart 104:1921–1928

Torres-Soto J, Ashley EA (2020) Multi-task deep learning for cardiac rhythm detection in wearable devices. NPJ Digit Med 3:116

Pluymaekers NAHA, Hermans ANL, van der Velden RMJ, Gawałko M, den Uijl DW, Buskes S, Vernooy K, Crijns HJGM, Hendriks JM, Linz D (2021) Implementation of an on-demand app-based heart rate and rhythm monitoring infrastructure for the management of atrial fibrillation through teleconsultation: TeleCheck-AF. Europace 23:345–352

Yan BP, Lai WHS, Chan CKY, Au ACK, Freedman B, Poh YC, Poh M‑Z (2020) High-throughput, contact-free detection of atrial fibrillation from video with deep learning. JAMA Cardiol 5:105–107

Liu X, Yang X, Wang D, Wong A, Ma L, Li L (2021) VidAF: a motion-robust model for screening atrial fibrillation from facial videos. IEEE J Biomed Health Inform. https://doi.org/10.1109/JBHI.2021.3124967

Jiang F, Hong C, Cheng T, Wang H, Xu B, Zhang B (2021) Attention-based multi-scale features fusion for unobtrusive atrial fibrillation detection using ballistocardiogram signal. BioMed Eng OnLine 20:12

A‑Rhythmik GmbH (2021) Prevention of stroke and sudden cardiac death by recording of 1‑channel electrocardiograms. https://clinicaltrials.gov/ct2/show/NCT04637230. Accessed 12 Dec 2021

Bhavnani S (2020) Development of a novel convolution neural network for arrhythmia classification for shockable cardiac rhythms. https://clinicaltrials.gov/ct2/show/NCT03662802. Accessed 13 Dec 2021

Lin C (2021) Artificial intelligence-enable electrocardiograms system for detecting atrial fibrillation in hospital. https://clinicaltrials.gov/ct2/show/NCT05127460. Accessed 12 Dec 2021

Yonsei University (2021) Continuous EKG monitoring using S‑patch ex : prospective observational study (S-patch registry). https://clinicaltrials.gov/ct2/show/NCT05119725. Accessed 12 Dec 2021

The University of Hong Kong (2020) Protocol for rationale and design of home-based solUtion for remote atrial fibrillation screening to PrevenT RecUrrence StrOke (HUA-TUO AF trial): a randomized open-label study. https://clinicaltrials.gov/ct2/show/NCT04523649. Accessed 12 Dec 2021

University of Leicester (2021) Validation of 3D stroke models to differentiate stroke subtypes and predict source in acute ischaemic stroke. https://clinicaltrials.gov/ct2/show/NCT05055960. Accessed 13 Dec 2021

Bundy JD, Heckbert SR, Chen LY, Lloyd-Jones DM, Greenland P (2020) Evaluation of risk prediction models of atrial fibrillation (from the multi-ethnic study of atherosclerosis [MESA]). Am J Cardiol 125:55–62

Karri R, Kawai A, Thong YJ, Ramson DM, Perry LA, Segal R, Smith JA, Penny-Dimri JC (2021) Machine learning outperforms existing clinical scoring tools in the prediction of postoperative atrial fibrillation during intensive care unit admission after cardiac surgery. Heart Lung Circ 30:1929–1937

Nadarajah R, Wu J, Frangi AF, Hogg D, Cowan C, Gale C (2021) Predicting patient-level new-onset atrial fibrillation from population-based nationwide electronic health records: protocol of FIND-AF for developing a precision medicine prediction model using artificial intelligence. BMJ Open 11:e52887

Khurshid S, Friedman S, Reeder C, Di Achille P, Diamant N, Singh P, Harrington LX, Wang X, Al-Alusi MA, Sarma G et al (2021) Electrocardiogram-based deep learning and clinical risk factors to predict atrial fibrillation. Circulation. https://doi.org/10.1161/CIRCULATIONAHA.121.057480

Rabinstein AA, Yost MD, Faust L, Kashou AH, Latif OS, Graff-Radford J, Attia IZ, Yao X, Noseworthy PA, Friedman PA (2021) Artificial intelligence-enabled ECG to identify silent atrial fibrillation in embolic stroke of unknown source. J Stroke Cerebrovasc Dis 30:105998

Yao X (2021) Batch enrollment for an artificial intelligence-guided intervention to lower neurologic events in patients with unrecognized atrial fibrillation (BEAGLE). https://clinicaltrials.gov/ct2/show/NCT04208971. Accessed 12 Dec 2021

Schwab K, Nguyen D, Ungab G, Feld G, Maisel AS, Than M, Joyce L, Peacock WF (2021) Artificial intelligence MacHIne learning for the detection and treatment of atrial fibrillation guidelines in the emergency department setting (AIM HIGHER): Assessing a machine learning clinical decision support tool to detect and treat non-valvular atrial fibrillation in the emergency department. J Am Coll Emerg Physicians Open 2:e12534

Levy AE, Biswas M, Weber R, Tarakji K, Chung M, Noseworthy PA, Newton-Cheh C, Rosenberg MA (2019) Applications of machine learning in decision analysis for dose management for dofetilide. PLoS ONE 14:e227324

Watanabe E, Noyama S, Kiyono K, Inoue H, Atarashi H, Okumura K, Yamashita T, Lip GYH, Kodani E, Origasa H (2021) Comparison among random forest, logistic regression, and existing clinical risk scores for predicting outcomes in patients with atrial fibrillation: a report from the J‑RHYTHM registry. Clin Cardiol 44:1305–1315

Samaras A, Kartas A, Akrivos E, Fotos G, Dividis G, Vasdeki D, Vrana E, Rampidis G, Karvounis H, Giannakoulas G et al (2021) A novel prognostic tool to predict mortality in patients with atrial fibrillation: the BASIC-AF risk score. Hellenic J Cardiol 62:339–348

Loring Z, Mehrotra S, Piccini JP, Camm J, Carlson D, Fonarow GC, Fox KAA, Peterson ED, Pieper K, Kakkar AK (2020) Machine learning does not improve upon traditional regression in predicting outcomes in atrial fibrillation: an analysis of the ORBIT-AF and GARFIELD-AF registries. Europace 22:1635–1644

Vinter N, Frederiksen AS, Albertsen AE, Lip GYH, Fenger-Grøn M, Trinquart L, Frost L, Møller DS (2020) Role for machine learning in sex-specific prediction of successful electrical cardioversion in atrial fibrillation? Open Heart 7:e1297

Shade JK, Ali RL, Basile D, Popescu D, Akhtar T, Marine JE, Spragg DD, Calkins H, Trayanova NA (2020) Preprocedure application of machine learning and mechanistic simulations predicts likelihood of paroxysmal atrial fibrillation recurrence following pulmonary vein isolation. Circ Arrhythm Electrophysiol 13:e8213

Yonsei University (2021) Artificial intelligence guided patient selection for atrial fibrillation catheter ablation: randomized clinical trial (AI-PAFA trial). https://clinicaltrials.gov/ct2/show/NCT04997824. Accessed 12 Dec 2021

Luongo G, Azzolin L, Schuler S, Rivolta MW, Almeida TP, Martínez JP, Soriano DC, Luik A, Müller-Edenborn B, Jadidi A et al (2021) Machine learning enables noninvasive prediction of atrial fibrillation driver location and acute pulmonary vein ablation success using the 12-lead ECG. Cardiovasc Digit Health J 2:126–136

Engdahl J (2021) Validation of a novel smartphone-based photoplethysmographic method for ambulatory heart rhythm monitoring in connection to treatment of atrial fibrillation with direct current cardioversion. https://clinicaltrials.gov/ct2/show/NCT04300270. Accessed 13 Dec 2021

Parikh RB, Teeple S, Navathe AS (2019) Addressing bias in artificial intelligence in health care. JAMA 322:2377–2378

Larson J, Mattu S, Kirchner L, Angwin J (2016) How we analyzed the COMPAS recidivism algorithm. https://www.propublica.org/article/how-we-analyzed-the-compas-recidivism-algorithm. Accessed 20 Dec 2021

Obermeyer Z, Powers B, Vogeli C, Mullainathan S (2019) Dissecting racial bias in an algorithm used to manage the health of populations. Science 366:447–453

Finlayson SG, Subbaswamy A, Singh K, Bowers J, Kupke A, Zittrain J, Kohane IS, Saria S (2021) The clinician and dataset shift in artificial intelligence. N Engl J Med 385:283–286

Yu K‑H, Beam AL, Kohane IS (2018) Artificial intelligence in healthcare. Nat Biomed Eng 2:719–731