{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T14:10:56Z","timestamp":1783433456863,"version":"3.54.6"},"reference-count":67,"publisher":"AIP Publishing","issue":"11","funder":[{"DOI":"10.13039\/501100004329","name":"Slovenian Research Agency","doi-asserted-by":"crossref","award":["P1-0403 and J1-2457"],"award-info":[{"award-number":["P1-0403 and J1-2457"]}],"id":[{"id":"10.13039\/501100004329","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["pubs.aip.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2021,11,1]]},"abstract":"<jats:p>In this study, we aimed to detect paroxysmal atrial fibrillation episodes before they occur so that patients can take precautions before putting their and others\u2019 lives in potentially life-threatening danger. We used the atrial fibrillation prediction database, open data from PhysioNet, and assembled our process based on convolutional neural networks. Conventional heart rate variability features are calculated from time-domain measures, frequency-domain measures using power spectral density estimations, time-frequency-domain measures using wavelet transform, and nonlinear Poincar\u00e9 plot measures. In addition, we also applied an alternative heart rate normalization, which gave promising results only in a few studies, before calculating these heart rate variability features. We used these features directly and their normalized versions using min\u2013max normalization and z-score normalization methods. Thus, heart rate variability features extracted from six different combinations of these normalizations, in addition to no normalization cases, were applied to the convolutional neural network classifier. We tuned the classifiers\u2019 hyperparameters using 90% of feature sets and tested the classifiers\u2019 performances using 10% of feature sets. The proposed approach resulted in 87.76% accuracy, 91.30% precision, 80.04% recall, and 87.50% f1-score in heart rate variability with z-score feature normalization. When the heart rate normalization was also utilized, the suggested method gave 100% accuracy, 100% precision, 100% recall, and 100% f1-score in heart rate variability with z-score feature normalization. The proposed method with heart rate normalization and z-score normalization methods resulted in better classification performance than similar studies in the literature. By comparing the existing studies, we conclude that our approach provides a much better tool to determine a near-future paroxysmal atrial fibrillation episode. However, although the achieved benchmarks are impressive, we note that the approach needs to be supported by other studies and on other datasets before clinical trials.<\/jats:p>","DOI":"10.1063\/5.0069272","type":"journal-article","created":{"date-parts":[[2021,11,9]],"date-time":"2021-11-09T18:02:18Z","timestamp":1636480938000},"update-policy":"https:\/\/doi.org\/10.1063\/aip-crossmark-policy-page","source":"Crossref","is-referenced-by-count":40,"title":["Convolutional neural networks predict the onset of paroxysmal atrial fibrillation: Theory and applications"],"prefix":"10.1063","volume":"31","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9889-9952","authenticated-orcid":false,"given":"M.","family":"Surucu","sequence":"first","affiliation":[{"name":"Department of Computer Engineering, Duzce University 1 , 81620 Duzce, Turkey"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2150-4756","authenticated-orcid":false,"given":"Y.","family":"Isler","sequence":"additional","affiliation":[{"name":"Department of Biomedical Engineering, Izmir Katip Celebi University 2 , Cigli, 35620 Izmir, Turkey"},{"name":"Islerya Medical and Information Technologies Company 3 , Bornova, 35030 Izmir, Turkey"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3087-541X","authenticated-orcid":false,"given":"M.","family":"Perc","sequence":"additional","affiliation":[{"name":"Faculty of Natural Sciences and Mathematics, University of Maribor 4 , Koro\u0161ka cesta 160, 2000 Maribor, Slovenia"},{"name":"Department of Medical Research, China Medical University Hospital, China Medical University 5 , Taichung 404332, Taiwan"},{"name":"Alma Mater Europaea 6 , Slovenska ulica 17, 2000 Maribor, Slovenia"},{"name":"Complexity Science Hub Vienna 7 , Josefst\u00e4dterstra\u00dfe 39, 1080 Vienna, Austria"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8902-6837","authenticated-orcid":false,"given":"R.","family":"Kara","sequence":"additional","affiliation":[{"name":"Department of Computer Engineering, Duzce University 1 , 81620 Duzce, Turkey"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"317","published-online":{"date-parts":[[2021,11,9]]},"reference":[{"key":"2023070318305535100_c1","doi-asserted-by":"publisher","first-page":"131","DOI":"10.1016\/S0002-9149(99)00713-4","article-title":"Epidemiology and significance of atrial fibrillation","volume":"84","year":"1999","journal-title":"Am. J. Cardiol."},{"key":"2023070318305535100_c2","doi-asserted-by":"publisher","first-page":"535","DOI":"10.1136\/heartjnl-2020-317915","article-title":"Atrial fibrillation prevalence, awareness and management in a nationwide survey of adults in China","volume":"107","year":"2021","journal-title":"Heart"},{"key":"2023070318305535100_c3","doi-asserted-by":"publisher","volume-title":"PAF Prediction Challenge Database","DOI":"10.13026\/C2H59W"},{"key":"2023070318305535100_c4","doi-asserted-by":"publisher","first-page":"373","DOI":"10.1093\/eurheartj\/ehaa612","article-title":"2020 ESC guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS)","volume":"42","year":"2020","journal-title":"Eur. Heart J."},{"key":"2023070318305535100_c5","doi-asserted-by":"publisher","first-page":"40","DOI":"10.1016\/j.cmpb.2010.07.011","article-title":"Prediction of paroxysmal atrial fibrillation based on non-linear analysis and spectrum and bispectrum features of the heart rate variability signal","volume":"105","year":"2012","journal-title":"Comput. Methods Programs Biomed."},{"key":"2023070318305535100_c6","doi-asserted-by":"publisher","first-page":"1502","DOI":"10.1016\/j.compbiomed.2007.01.012","article-title":"Combining classical HRV indices with wavelet entropy measures improves to performance in diagnosing congestive heart failure","volume":"37","year":"2007","journal-title":"Comput. Biol. Med."},{"key":"2023070318305535100_c7","doi-asserted-by":"publisher","first-page":"472","DOI":"10.1016\/j.procs.2018.04.052","article-title":"Atrial fibrillation detection based on ECG-features extraction in WBSN","volume":"130","year":"2018","journal-title":"Procedia Comput. Sci."},{"key":"2023070318305535100_c8","doi-asserted-by":"publisher","first-page":"62","DOI":"10.1016\/j.knosys.2017.06.003","article-title":"Automated detection of coronary artery disease using different durations of ECG segments with convolutional neural network","volume":"132","year":"2017","journal-title":"Knowl. Based Syst."},{"key":"2023070318305535100_c9","doi-asserted-by":"publisher","first-page":"465","DOI":"10.3906\/elk-1207-132","article-title":"Feature selection on single-lead ECG for obstructive sleep apnea diagnosis","volume":"22","year":"2014","journal-title":"Turk. J. Electr. Eng. Comput. Sci."},{"key":"2023070318305535100_c10","doi-asserted-by":"publisher","first-page":"1043","DOI":"10.1161\/01.CIR.93.5.1043","article-title":"Heart rate variability","volume":"93","author":"Task Force of the European Society of Cardiology the North American Society of Pacing Electrophysiology","year":"1996","journal-title":"Circulation"},{"key":"2023070318305535100_c11","doi-asserted-by":"publisher","first-page":"95","DOI":"10.1016\/S0169-2607(03)00079-8","article-title":"Automatic arrhythmia detection based on time and time\u2013frequency analysis of heart rate variability","volume":"74","year":"2004","journal-title":"Comput. Methods Programs Biomed."},{"key":"2023070318305535100_c12"},{"key":"2023070318305535100_c13","doi-asserted-by":"publisher","first-page":"38","DOI":"10.1186\/1475-925X-8-38","article-title":"Atrial fibrillation detection by heart rate variability in Poincar\u00e9 plot","volume":"8","year":"2009","journal-title":"Biomed. Eng. Online"},{"key":"2023070318305535100_c14","doi-asserted-by":"publisher","first-page":"113","DOI":"10.1016\/j.compbiomed.2016.06.029","article-title":"Discrimination of systolic and diastolic dysfunctions using multi-layer perceptron in heart rate variability analysis","volume":"76","year":"2016","journal-title":"Comput. Biol. Med."},{"key":"2023070318305535100_c15","doi-asserted-by":"publisher","first-page":"56","DOI":"10.1016\/j.physa.2018.06.022","article-title":"Early prediction of paroxysmal atrial fibrillation based on short-term heart rate variability","volume":"509","year":"2018","journal-title":"Physica A"},{"key":"2023070318305535100_c16","doi-asserted-by":"publisher","first-page":"453","DOI":"10.1243\/09544119JEIM642","article-title":"Heart rate normalization in the analysis of heart rate variability in congestive heart failure","volume":"224","year":"2009","journal-title":"Proc. Inst. Mech. Eng., Part H: J. Eng. Med."},{"key":"2023070318305535100_c17"},{"key":"2023070318305535100_c18"},{"key":"2023070318305535100_c19","doi-asserted-by":"publisher","first-page":"187","DOI":"10.1016\/j.cmpb.2016.07.016","article-title":"Paroxysmal atrial fibrillation prediction method with shorter HRV sequences","volume":"134","year":"2016","journal-title":"Comput. Methods Programs Biomed."},{"key":"2023070318305535100_c20"},{"key":"2023070318305535100_c21","doi-asserted-by":"publisher","first-page":"561","DOI":"10.1109\/TBME.2003.821030","article-title":"Prediction of paroxysmal atrial fibrillation by analysis of atrial premature complexes","volume":"51","year":"2004","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"2023070318305535100_c22","doi-asserted-by":"publisher","first-page":"679","DOI":"10.1016\/j.compbiomed.2003.10.002","article-title":"ECG characterization of paroxysmal atrial fibrillation: Parameter extraction and automatic diagnosis algorithm","volume":"34","year":"2004","journal-title":"Comput. Biol. Med."},{"key":"2023070318305535100_c23"},{"key":"2023070318305535100_c24"},{"key":"2023070318305535100_c25","doi-asserted-by":"publisher","first-page":"1959","DOI":"10.1088\/0967-3334\/33\/12\/1959","article-title":"Study on the p-wave feature time course as early predictors of paroxysmal atrial fibrillation","volume":"33","year":"2012","journal-title":"Physiol. Meas."},{"key":"2023070318305535100_c26","doi-asserted-by":"publisher","first-page":"2095","DOI":"10.1109\/TSMC.2017.2705582","article-title":"Deep convolutional neural networks and learning ECG features for screening paroxysmal atrial fibrillation patients","volume":"48","year":"2018","journal-title":"IEEE Trans. Syst., Man, Cybern.: Syst."},{"key":"2023070318305535100_c27","doi-asserted-by":"publisher","first-page":"2986","DOI":"10.3906\/elk-2105-92","article-title":"Diagnosis of paroxysmal atrial fibrillation from thirty-minute heart rate variability data using convolutional neural networks","volume":"29","year":"2021","journal-title":"Turk. J. Electr. Eng. Comput. Sci."},{"key":"2023070318305535100_c28"},{"key":"2023070318305535100_c29"},{"key":"2023070318305535100_c30","doi-asserted-by":"publisher","first-page":"151","DOI":"10.1016\/j.artmed.2008.03.009","article-title":"Complexity and spectral analysis of the heart rate variability dynamics for distant prediction of paroxysmal atrial fibrillation with artificial intelligence methods","volume":"43","year":"2008","journal-title":"Artif. Intell. Med."},{"key":"2023070318305535100_c31","doi-asserted-by":"publisher","first-page":"171","DOI":"10.1016\/j.cmpb.2017.10.012","article-title":"Paroxysmal atrial fibrillation prediction based on HRV analysis and non-dominated sorting genetic algorithm III","volume":"153","year":"2018","journal-title":"Comput. Methods Programs Biomed."},{"key":"2023070318305535100_c32"},{"key":"2023070318305535100_c33","first-page":"134","article-title":"Effect of deep learning feature inference techniques on respiratory sounds","volume":"3","year":"2020","journal-title":"J. Intell. Syst. Appl."},{"key":"2023070318305535100_c34","first-page":"130","article-title":"Examining the difference between image size, background color, gray picture and color picture in leave classification with deep learning","volume":"3","year":"2020","journal-title":"J. Intell. Syst. Appl."},{"key":"2023070318305535100_c35","doi-asserted-by":"publisher","first-page":"58","DOI":"10.1016\/j.bspc.2018.05.014","article-title":"Deep learning with 3D-second order difference plot on respiratory sounds","volume":"45","year":"2018","journal-title":"Biomed. Signal Process. Control"},{"key":"2023070318305535100_c36","doi-asserted-by":"publisher","first-page":"1344","DOI":"10.1109\/JBHI.2019.2931395","article-title":"Deep learning on computerized analysis of chronic obstructive pulmonary disease","volume":"24","year":"2020","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"2023070318305535100_c37"},{"key":"2023070318305535100_c38","doi-asserted-by":"publisher","first-page":"41","DOI":"10.1023\/A:1026507317626","article-title":"Validity test for a set of nonlinear measures for short data length with reference to short-term heart rate variability signal","volume":"10","year":"2000","journal-title":"J. Syst. Integr."},{"key":"2023070318305535100_c39","doi-asserted-by":"publisher","first-page":"230","DOI":"10.1109\/TBME.1985.325532","article-title":"A real-time QRS detection algorithm","volume":"BME-32","year":"1985","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"2023070318305535100_c40"},{"key":"2023070318305535100_c41","doi-asserted-by":"publisher","first-page":"1414","DOI":"10.1109\/TBME.2004.828049","article-title":"Structural relationships between measures based on heart beat intervals: Potential for improved risk assessment","volume":"51","year":"2004","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"2023070318305535100_c42","doi-asserted-by":"publisher","first-page":"623","DOI":"10.1111\/j.1469-8986.1997.tb02140.x","article-title":"Heart rate variability: Origins, methods, and interpretive caveats","volume":"34","year":"1997","journal-title":"Psychophysiology"},{"key":"2023070318305535100_c43","doi-asserted-by":"publisher","first-page":"630","DOI":"10.1109\/TBME.2005.844028","article-title":"Quantifying errors in spectral estimates of HRV due to beat replacement and resampling","volume":"52","year":"2005","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"2023070318305535100_c44","volume-title":"Advanced Methods and Tools for ECG Data Analysis","year":"2006"},{"key":"2023070318305535100_c45","doi-asserted-by":"publisher","first-page":"55","DOI":"10.1111\/j.1469-8986.1992.tb02011.x","article-title":"A nonstationarity test for the spectral analysis of physiological time series with an application to respiratory sinus arrhythmia","volume":"29","year":"1992","journal-title":"Psychophysiology"},{"key":"2023070318305535100_c46","doi-asserted-by":"publisher","first-page":"172","DOI":"10.1109\/10.979357","article-title":"An advanced detrending method with application to HRV analysis","volume":"49","year":"2002","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"2023070318305535100_c47","volume-title":"Pattern Classification","year":"2001","edition":"2nd ed."},{"key":"2023070318305535100_c48","doi-asserted-by":"publisher","first-page":"447","DOI":"10.1007\/BF00648343","article-title":"Least-squares frequency analysis of unequally spaced data","volume":"39","year":"1976","journal-title":"Astrophys. Space Sci."},{"key":"2023070318305535100_c49","doi-asserted-by":"publisher","first-page":"835","DOI":"10.1086\/160554","article-title":"Studies in astronomical time series analysis. II. Statistical aspects of spectral analysis of unevenly spaced data","volume":"263","year":"1982","journal-title":"Astrophys. J."},{"key":"2023070318305535100_c50","doi-asserted-by":"publisher","first-page":"698","DOI":"10.1109\/10.678605","article-title":"Power spectral density of unevenly sampled data by least-square analysis: Performance and application to heart rate signals","volume":"45","year":"1998","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"2023070318305535100_c51","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1111\/j.1445-5994.1995.tb00573.x","article-title":"Application of the Poincar\u00e9 plot to heart rate variability: A new measure of functional status in heart failure","volume":"25","year":"1995","journal-title":"Aust. N. Z. J. Med."},{"key":"2023070318305535100_c52","doi-asserted-by":"publisher","first-page":"201","DOI":"10.1042\/cs0910201","article-title":"Poincar\u00e9 plot of heart rate variability allows quantitative display of parasympathetic nervous activity in humans","volume":"91","year":"1996","journal-title":"Clin. Sci."},{"key":"2023070318305535100_c53","doi-asserted-by":"publisher","first-page":"1342","DOI":"10.1109\/10.959330","article-title":"Do existing measures of Poincar\u00e9 plot geometry reflect nonlinear features of heart rate variability?","volume":"48","year":"2001","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"2023070318305535100_c54"},{"key":"2023070318305535100_c55","doi-asserted-by":"publisher","first-page":"291","DOI":"10.1007\/s004220000212","article-title":"Wavelet entropy in event-related potentials: A new method shows ordering of EEG oscillations","volume":"84","year":"2001","journal-title":"Biol. Cybern."},{"key":"2023070318305535100_c56","doi-asserted-by":"publisher","first-page":"593","DOI":"10.17341\/gazimmfd.664583","article-title":"Evaluation of wavelet features selected via statistical evidence from steady-state visually evoked potentials to predict the stimulating frequency","volume":"36","year":"2021","journal-title":"J. Fac. Eng. Archit. Gazi Univ."},{"key":"2023070318305535100_c57","doi-asserted-by":"publisher","first-page":"2263","DOI":"10.3906\/elk-2010-26","article-title":"Evaluation of mother wavelets on steady-state visually-evoked potentials for triple-command brain-computer interfaces","volume":"29","year":"2021","journal-title":"Turk. J. Electr. Eng. Comput. Sci."},{"key":"2023070318305535100_c58","doi-asserted-by":"publisher","first-page":"85","DOI":"10.1016\/j.neunet.2014.09.003","article-title":"Deep learning in neural networks: An overview","volume":"61","year":"2015","journal-title":"Neural Netw."},{"key":"2023070318305535100_c59","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","year":"2015","journal-title":"Nature"},{"key":"2023070318305535100_c60","doi-asserted-by":"publisher","first-page":"1798","DOI":"10.1109\/TPAMI.2013.50","article-title":"Representation learning: A review and new perspectives","volume":"35","year":"2013","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"2023070318305535100_c61","first-page":"1929","article-title":"Dropout: A simple way to prevent neural networks from overfitting","volume":"15","year":"2014","journal-title":"J. Mach. Learn. Res."},{"key":"2023070318305535100_c62","first-page":"281","article-title":"Random search for hyper-parameter optimization","volume":"13","year":"2012","journal-title":"J. Mach. Learn. Res."},{"key":"2023070318305535100_c63","doi-asserted-by":"publisher","first-page":"1221","DOI":"10.1109\/TMI.2017.2690836","article-title":"Automatic quality assessment of echocardiograms using convolutional neural networks: Feasibility on the apical four-chamber view","volume":"36","year":"2017","journal-title":"IEEE Trans. Med. Imaging"},{"key":"2023070318305535100_c64","doi-asserted-by":"publisher","first-page":"196","DOI":"10.1515\/msr-2015-0027","article-title":"Comparison of the effects of cross-validation methods on determining performances of classifiers used in diagnosing congestive heart failure","volume":"15","year":"2015","journal-title":"Meas. Sci. Rev."},{"key":"2023070318305535100_c65","doi-asserted-by":"publisher","first-page":"6","DOI":"10.1186\/s12864-019-6413-7","article-title":"The advantages of the Matthews correlation coefficient (MCC) over f1 score and accuracy in binary classification evaluation","volume":"21","year":"2020","journal-title":"BMC Genom."},{"key":"2023070318305535100_c66"},{"key":"2023070318305535100_c67"}],"container-title":["Chaos: An Interdisciplinary Journal of Nonlinear Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.aip.org\/aip\/cha\/article-pdf\/doi\/10.1063\/5.0069272\/14665574\/113119_1_online.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.aip.org\/aip\/cha\/article-pdf\/doi\/10.1063\/5.0069272\/14665574\/113119_1_online.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,7,3]],"date-time":"2023-07-03T21:38:11Z","timestamp":1688420291000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.aip.org\/cha\/article\/31\/11\/113119\/342158\/Convolutional-neural-networks-predict-the-onset-of"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,11,1]]},"references-count":67,"journal-issue":{"issue":"11","published-print":{"date-parts":[[2021,11,1]]}},"URL":"https:\/\/doi.org\/10.1063\/5.0069272","relation":{},"ISSN":["1054-1500","1089-7682"],"issn-type":[{"value":"1054-1500","type":"print"},{"value":"1089-7682","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2021,11]]},"published":{"date-parts":[[2021,11,1]]},"article-number":"113119"}}