Diagnosis of paroxysmal atrial fibrillation from thirty-minute heart rate variability data using convolutional neural networks

Paroxysmal atrial fibrillation (PAF) is the initial stage of atrial fibrillation, one of the most common arrhythmia types. PAF worsens with time and affects the patient’s life quality negatively. In this study, we aimed to diagnose PAF early, so patients can start taking precautions before this disease gets worse. We used the atrial fibrillation prediction database, an open data from Physionet and constructed our approach using convolutional neural networks. Heart rate variability (HRV) features are calculated from time-domain measures, frequency-domain measures using power spectral density estimations (fast Fourier transform, Lomb–Scargle, and Welch periodogram), time-frequencydomain measures using wavelet transform, and nonlinear Poincare plot measures. We also normalized these features using min-max normalization and z-score normalization methods. In addition, we also applied alternatively the heart rate normalization (HRN), which gave promising results in a few HRV-based research, before calculating these features. Thus, HRV data, HRN data, and HRV features extracted from six different combinations of these normalizations, in addition to no normalization cases, were applied to the convolutional neural networks classifier. We tuned the classifiers using 90% of samples and tested the classifiers’ performances using 10% of data. The proposed approach resulted in 95.92% accuracy, 100% precision, 91.84% recall, and 95.74% f1-score in HRV with z-score feature normalization. When the heart rate normalization was also applied, the proposed approach reached 100% accuracy, 100% precision, 100% recall, and 100% f1-score in HRV 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. In addition, although deep learning models offer no use of separate feature extraction processes, this study reveals that using HRV-specific feature extraction techniques may improve the performance of deep learning algorithms in HRV-based studies. Comparing the existing studies, we concluded that our approach provides a much better tool to diagnose PAF patients.

PDF

___

[1] Ryder KM, Benjamin EJ. Epidemiology and significance of atrial fibrillation. The American Journal of Cardiology 1999; 84 (9A): 131R-138R. doi: 10.1016/s0002-9149(99)00713-4
[2] Mohebbi M, Ghassemian H. Prediction of paroxysmal atrial fibrillation based on non-linear analysis and spectrum and bispectrum features of the heart rate variability signal. Computer Methods and Programs in Biomedicine 2012; 105 (1): 40-49. doi: 10.1016/j.cmpb.2010.07.011
[3] Isler Y, Kuntalp M. Combining classical hrv indices with wavelet entropy measures improves to performance in diagnosing congestive heart failure, Computers in Biology and Medicine 2007; 37 (10): 1502-1510. doi: 10.1016/j.compbiomed.2007.01.012
[4] Gladstone DJ, Spring M, Dorian P, Panzov V, Thorpe KE et al. Atrial fibrillation in patients with cryptogenic stroke. New England Journal of Medicine 2014; 370 (26): 2467-2477. doi: 10.1056/nejmoa1311376
[5] Narin A, Isler Y, Ozer M, Perc M. Early prediction of paroxysmal atrial fibrillation based on short-term heart rate variability. Physica A: Statistical Mechanics and its Applications 2018; 509: 56-65. doi: 10.1016/j.physa.2018.06.022
[6] Soudani A, Almusallam M. Atrial fibrillation detection based on ECG-Features extraction in WBSN. Procedia Computer Science 2018; 130: 472-479. doi: 10.1016/j.procs.2018.04.052
[7] Acharya UR, Fujita H, Lih OS, Adam M, Tan JH et al. Automated detection of coronary artery disease using different durations of ECG segments with convolutional neural network. Knowledge-Based Systems 2017; 132: 62-71. doi: 10.1016/j.knosys.2017.06.003
[8] Guruler H, Sahin M, Ferikoglu A. Feature selection on single-lead ECG for obstructive sleep apnea diagnosis. Turkish Journal of Electrical Engineering and Computer Science 2014; 22 (2): 465-478. doi: 10.3906/elk-1207-132
[9] Camm A, Malik JM, Bigger J, Brethardt TG, Cerutti S et al. Heart rate variability. Circulation 1996; 93 (5): 1043-1065. doi: 10.1161/01.CIR.93.5.1043
[10] Tsipouras MG, Fotiadis DI. Automatic arrhythmia detection based on time and time-frequency analysis of heart rate variability. Computer Methods and Programs in Biomedicine 2004; 74 (2): 95-108. doi: 10.1016/S0169- 2607(03)00079-8
[11] Ming-Yuan L, Sung-Nien Y. Multiscale sample entropy based on discrete wavelet transform for clinical heart rate variability recognition. Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) 2012; 4299-4302. doi: 10.1109/EMBC.2012.6346917
[12] Park J, Lee S, Jeon M. Atrial fibrillation detection by heart rate variability in Poincare plot. BioMedical Engineering OnLine 2009; 8 (1): 38. doi: 10.1186/1475-925X-8-38
[13] Seker R, Saliu S, Birand A, Kudaiberdieva G. Validity test for a set of nonlinear measures for short data length with reference to short-term heart rate variability signal. Journal of Systems Integration 2000; 10: 41-53. doi: 10.1023/A:1026507317626
[14] Moody GB, Goldberger AL, McClennen S, Swiryn SP. Predicting the onset of paroxysmal atrial fibrillation: The computers in cardiology challenge. Computers in Cardiology 2001; 28: 113-116. doi: 10.1109/CIC.2001.977604
[15] Hallstrom AP, Stein PK, Schneider R, Hodges M, Schmidt G et al. Structural relationships between measures based on heart beat intervals: potential for improved risk assessment. IEEE Transactions on Biomedical Engineering 2004; 51 (8): 1414-1420. doi: 10.1109/TBME.2004.828049
[16] Isler Y. Discrimination of Systolic and Diastolic Dysfunctions using Multi-Layer Perceptron in Heart Rate Variability Analysis. Computers in Biology and Medicine 2016; 76: 113-119. doi: 10.1016/j.compbiomed.2016.06.029
[17] Surucu M, Isler Y, Kara R. Investigation of the effect of normalization techniques on discriminating patients with paroxysmal atrial fibrillation. In: 2nd International Conference of Applied Sciences, Engineering and Mathematics (IBU-ICASEM 2020), June 4-6, Skopje/North Macedonia, Book of Abstracts, 2020, ISBN: 978-608-4868-02-6, 12.
[18] Isler Y, Kuntalp M. Heart rate normalization in the analysis of heart rate variability in congestive heart failure. Proceedings of the IMechE Part H: Journal of Engineering in Medicine 2010; 224 (3): 453–463. doi: 10.1243/09544119JEIM642
[19] Maier C, Bauch M, Dickhaus H. Screening and prediction of paroxysmal atrial fibrillation by analysis of heart rate variability parameters. Computers in Cardiology 2001; 28: 129-132. doi: 10.1109/cic.2001.977608
[20] Schreier G, Kastner P, Marko W. An automatic ECG processing algorithm to identify patients prone to paroxysmal atrial fibrillation. Computers in Cardiology 2001; 28: 133-135. doi: 10.1109/cic.2001.977609
[21] Boon KH, Khalil-Hani M, Malarvili MB. Paroxysmal atrial fibrillation prediction based on HRV analysis and non-dominated sorting genetic algorithm. Computer Methods and Programs Biomedicine 2018; 153: 171-184. doi: 10.1016/j.cmpb.2017.10.012
[22] Boon KH, Khalil-Hani M, Malarvili MB, Sia CW. Paroxysmal atrial fibrillation prediction method with shorter HRV sequences. Computer Methods and Programs Biomedicine 2016; 134: 187-196. doi: 10.1016/j.cmpb.2016.07.016
[23] Thong T, McNames J, Aboy M, Goldstein B. Prediction of paroxysmal atrial fibrillation by analysis of atrial premature complexes. IEEE Transactions on Biomedical Engineering 2004; 51: 561-569. doi: 10.1109/TBME.2003.821030
[24] Ros E, Mota S, Fernandez FJ, Toro FJ, Bernier JL. ECG characterization of paroxysmal atrial fibrillation: parameter extraction and automatic diagnosis algorithm. Computers in Biology and Medicine 2004; 34 (8): 679-696. doi: 10.1016/j.compbiomed.2003.10.002
[25] Ozcan N, Kuntalp M. Determining best HRV indices for PAF screening using genetic algorithm. 10th International Conference on Electrical and Electronics Engineering (ELECO), 2017; 1385-1388.
[26] Chesnokov YV, Holden AV, Zhang H. Screening patients with paroxysmal atrial fibrillation (PAF) from non-PAF heart rhythm using HRV data analysis. Computers in Cardiology 2007; 2007: 459–462. doi: 10.1109/cic.2007.4745521
[27] Martinez A, Alcaraz R, Rieta JJ. Study on the P-wave feature time course as early predictors of paroxysmal atrial fibrillation. Physiological Measurement 2012; 33 (12): 1959-1974. doi: 10.1088/0967-3334/33/12/1959
[28] Balli O, Kutlu Y. Effect of deep learning feature inference techniques on respiratory sounds. Journal of Intelligent Systems with Applications 2020; 3 (2): 134-137 (in Turkish with an abstract in English). http://asd.islerya. com/indir.php?id=135
[29] Camgozlu Y, Kutlu Y. Examining the difference between image size, background color, gray picture and color picture in leave classification with deep learning. Journal of Intelligent Systems with Applications 2020; 3 (2): 130-133 (in Turkish with an abstract in English). http://asd.islerya.com/indir.php?id=133
[30] Altan G, Kutlu, Y, Pekmezci AO, Nural S. Deep learning with 3D-second order difference plot on respiratory sound. Biomedical Signal Processing and Control 2018; 45: 58-69. doi: 10.1016/j.bspc.2018.05.014
[31] Altan G, Kutlu Y, Allahverdi N. Deep learning on computerized analysis of chronic obstructive pulmonary disease. IEEE Journal of Biomedical and Health Informatics 2020; 24 (5): 1344-1350. doi: 10.1109/JBHI.2019.2931395
[32] Pourbabaee B, Roshtkhari MJ, Khorasani K. Deep convolutional neural networks and learning ecg features for screening paroxysmal atrial fibrillation patients. IEEE Transactions on Systems, Man, and Cybernetics: Systems 2018; 48 (12): 2095-2104. doi: 10.1109/tsmc.2017.2705582
[33] Pan J, Tompkins WJ. A real–time QRS detection algorithm. IEEE Transactions on Biomedical Engineering 1985; 32 (3): 230-236. doi: 10.1109/TBME.1985.325532
[34] Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation 1996; 93: 1043-1065. doi: 10.1161/01.CIR.93.5.1043
[35] Langley P, Di Bernardo D, Allen J, Bowers E, Smith F et al. Can paroxysmal atrial fibrillation be predicted? Computers in Cardiology 2001; 28: 121-124. doi: 10.1109/CIC.2001.977606
[36] Berntson G, Bigger J, Eckberg D, Grossman P, Kaufmann P et al. Heart rate variability: origins, methods, and interpretive caveats. Psychophysiology 1997; 34: 623-648. doi: 10.1111/j.1469-8986.1997.tb02140.x
[37] Clifford GD, Tarassenko L. Quantifying errors in spectral estimates of HRV due to beat replacement and resampling. IEEE Transactions on Biomedical Engineering 2005; 52 (4): 630-638. 10.1109/TBME.2005.844028
[38] Clifford GD, Azuaje F, McSharry PE. Advanced Methods and Tools for ECG Data Analysis. Norwood, MA, USA: Artech House, 2006.
[39] Weber EJM, Molenaar P, Molen MW. A nonstationarity test for the spectral analysis of physiological time series with an application to respiratory sinus arrhythmia. Psychophysiology 1992; 29 (1): 55-65. doi: 10.1111/j.1469- 8986.1992.tb02011.x
[40] Tarvainen MP, Ranta-Aho PO, Karjalainen PA. An advanced detrending method with application to HRV analysis. IEEE Transactions on Biomedical Engineering 2002; 49 (2): 172-175. doi: 10.1109/10.979357
[41] Duda RO, Hart PE, Stork DG. Pattern Classification. 2nd edition, New York, NY, USA: Wiley-Interscience, 2000.
[42] Lomb NR. Least-squares frequency analysis of unequally space data. Astrophysics and Space Science 1976; 39 (2): 447–462. doi: 10.1007/BF00648343
[43] Scargle JD. Studies in astronomical time series analysis (II): statistical aspects of spectral analysis of unevenly spaced data. Astrophysical Journal 1982; 263: 835-853. doi: 10.1086/160554
[44] Laguna P, Moody GB, Mark RG. Power spectral density of unevenly sampled data by least square analysis: performance and application to heart rate signals. IEEE Transactions on Biomedical Engineering 1998; 45 (6): 698-715. doi: 10.1109/10.678605
[45] Kamen PW, Tonkin AM. Application of the Poincare plot to heart rate variability: a new measure of functional status in heart failure. Australian and New Zealand Journal of Medicine 1995; 25 (1): 18–26. doi: 10.1111/j.1445- 5994.1995.tb00573.x
[46] Kamen PW, Krum H, Tonkin AM. Poincare plot of heart rate variability allows quantitative display of parasympathetic nervous activity. Clinical Science (London, England) 1996; 92: 201–208. doi: 10.1042/cs0910201
[47] Brennan M, Palaniswami M, Kamen P. Do existing measures of Poincare plot geometry reflect nonlinear features of heart rate variability? IEEE Transactions on Biomedical Engineering 2001; 48 (11): 1342-1347. doi: 10.1109/10.959330
[48] Marciano ML, Migaux F, Acanfora D, Furgi G, Rengo F. Quantification of Poincaree maps for the evaluation of heart rate variability. Computers in Cardiology 1994; 1994: 557–580. doi: 10.1109/CIC.1994.470126
[49] Quiraga RQ, Rosso OA, Basar E, Schurmann M. Wavelet entropy in event-related potentials: a new method shows ordering of EEG oscillations. Biological Cybernetics 2001; 84 (4): 291-299. doi: 10.1007/s004220000212
[50] Sayilgan E, Yuce YK, Isler Y. Evaluation of wavelet features selected via statistical evidence from steady-state visually-evoked potentials to predict the stimulating frequency. Journal of the Faculty of Engineering and Architecture of Gazi University 2021; 36 (2): 593-605. doi: 10.17341/gazimmfd.664583
[51] Sayilgan E, Yuce YK, Isler Y. Evaluation of mother wavelets on steady-state visually-evoked potentials for triplecommand brain-computer interfaces. Turkish Journal of Electrical Engineering & Computer Sciences 2021; in press. doi: 10.3906/elk-2010-26
[52] Schmidhuber J. Deep learning in neural networks: An overview. Neural Networks 2015; 61: 85-117. doi: 10.1016/j.neunet.2014.09.003
[53] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015; 521 (7553): 436-444. doi: 10.1038/nature14539
[54] Bengio Y, Courville A, Vincent P. Representation learning: A review and new perspectives. IEEE Transactions on Pattern Analysis and Machine Intelligence 2013; 35 (8): 1798-1828. doi: 10.1109/tpami.2013.50
[55] Chollet F. Deep Learning with Python. 2nd edition, Shelter Island, NY, USA: Manning Publications, 2020.
[56] Srivastava N, Hinton G, Krizhevsky A, Sutskever A, Salakhutdinov R. Dropout: A simple way to prevent neural networks from overfitting. Journal of Machine Learning Research 2014; 15 (56): 1929- 1958.https://jmlr.org/papers/volume15/srivastava14a/srivastava14a.pdf
[57] Kingma DP, Ba JL. Adam: a method for stochastic optimization. 3rd International Conference for Learning Representations, San Diego, 2015. https://arxiv.org/abs/1412.6980v9
[58] Bergstra J, Bengio Y. Random search for hyper-parameter optimization. Journal of Machine Learning Research 2012; 13: 281-305. https://www.jmlr.org/papers/volume13/bergstra12a/bergstra12a.pdf
[59] Abdi AH, Luong C, Tsang T, Allan G, Nouranian S et al. Automatic quality assessment of echocardiograms using convolutional neural networks: feasibility on the apical four-chamber view. IEEE Transactions on Medical Imaging 2017; 36 (6): 1221-1230. doi: doi: 10.1109/TMI.2017.2690836
[60] Chicco D, Jurman G. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genomics 2020; 21: 1-6. doi: 10.1186/s12864-019-6413-7