طبقه بندی آریتمی انقباضات زودرس بطنی از روی سیگنال ECG به کمک استخراج ویژگی های چند حوزه ای و کاهش ابعاد مبتنی بر کدگذار خودکار متغیر
مریم ابراهیم پورکازرونی
1
(
گروه مهندسی برق و کامپیوتر، واحد کازرون، دانشگاه آزاد اسلامی، کازرون، ایران
)
مهدی تقی زاده
2
(
گروه مهندسی برق و کامپیوتر، واحد کازرون، دانشگاه آزاد اسلامی، کازرون، ایران
)
محمد حسین فاتحی دیندارلو
3
(
گروه مهندسی برق و کامپیوتر، واحد کازرون، دانشگاه آزاد اسلامی، کازرون، ایران
)
امید مهدیار
4
(
گروه مهندسی برق و کامپیوتر، واحد کازرون، دانشگاه آزاد اسلامی، کازرون، ایران
)
جاسم جمالی
5
(
گروه مهندسی برق و کامپیوتر، واحد کازرون، دانشگاه آزاد اسلامی، کازرون، ایران
)
کلید واژه: انقباضات زودرس بطنی, الکتروکاردیوگرام, استخراج ویژگی, کدگذار خودکار, کاهش ویژگی, یادگیری ماشین,
چکیده مقاله :
یکی از خطرناکترین بیماریهای قلبی و عامل اصلی مرگ و میر در سطح جهانی، اختلالات قلبی عروقی می باشد. یکی از رایجترین انواع آریتمی قلبی، انقباضات زودرس بطنی (PVC) است. ثبت و تحلیل سیگنالهای الکتروکاردیوگرام، یکی از روشهای رایج، غیرتهاجمی وکم هزینه برای بررسی اختلالات قلبی است. این پژوهش، یک رویکرد جدید با نظارت برای طبقه بندی خودکار انقباضات زودرس بطنی را معرفی می کند که شامل ترکیبی از ویژگی های استخراج شده از سیگنال های الکتروکاردیوگرام و کاهش ویژگی ها بر اساس یادگیری عمیق است. روش پیشنهادی با بهره گیری از چندین تکنیک، از جمله تحلیل آماری، آشوب در حوزه زمان و حوزه زمان-فرکانس و ارزیابی ریخت شناسی به استخراج ویژگی‌ها از سیگنالهای الکتروکاردیوگرام پرداخته است. سپس، شبکه کدگذار متغیر خودکار، به عنوان یک روش پیشنهادی برای کاهش ویژگی، مورد بررسی قرارگرفته است. در انتها، برای طبقهبندی سیگنال الکتروکاردیوگرام، از ماشین بردار پشتیبان، k-نزدیک ترین همسایه و طبقه بند شبکه عصبی با استفاده از اعتبار سنجی k-fold بهرهگیری شده است. علاوه بر این پایگاه داده MIT-BIH به منظور ارزیابی روش پیشنهادی مورد استفاده گرفته است. نتایج حاصل نشان می دهد که روش پیشنهادی با استفاده از طبقه بند ماشین بردار پشتیبان به صحت 85/99درصد وحساسیت ۹5/99درصد و ویژه بودن 42/99درصد دست یافته است که این عملکرد نسبت به مطالعات مشابه موجود، بهبودیافته است.
چکیده انگلیسی :
The most serious disease and the main causes of death worldwide are recognized to be cardiovascular ailments.the most common kinds of cardiac arrhythmia is premature ventricular contraction (PVC). the most common (and least invasive and Low costs) techniques for examining cardiac problems is the recording and analysis of electrocardiogram (ECG) data. Through a combination of manually extracted features from ECG signals and feature reduction based on deep learning, a novel supervised strategy for the automatic identification of PVC has been established in this work. The suggested technique extracted several properties from ECG signals using a variety of approaches, including statistical,chaotic analysis in the time domain and time-frequency domain, as well as morphological evaluation. Then, in order to minimize the amount of extracted features and acquire the most discriminating features, a variational autoencoder (VAE) network is constructed as a deep learning based feature reduction approach. In order to identify ECG data, support vector machines, k-nearest neighbors, and neural network classifiers with five-fold cross-validation are used. The suggested strategy is assessed using the MIT-BIH database. The obtained results showed that the proposed method using SVM classifier reached 99/85% accuracy, 99/95% sensitivity, 99/42% specificity, which has performed better compared to existing similar studies
[1] R.Jagannathan, S. A. Patel, M. K. Ali and K. M. V. Narayan, "Global updates on cardiovascular disease mortality trends and attribution of traditional risk factors", Current diabetes reports, vol. 19, no. 44, pp. 44, 2019, doi: 10.1007/s11892-019-1161-2.
[2] J. Bernardi, K. A. Aromolaran, A. S Aromolaran " Neurological disorders and risk of arrhythmia ". International Journal of Molecular Sciences, vol. 22, no. 1, pp. 188, 2020, doi: 10.3390/ijms22010188.
[3] S. Sahoo, M. Dash, S. Behera, S. Sabut," Machine learning approach to detect cardiac arrhythmias in ECG signals: a survey", Irbm, vol. 41, no. 4, pp.185-194, 2020, doi: 10.1016/j.irbm.2019.12.001.
[4] H. Li, P. Boulanger, "A survey of heart anomaly detection using ambulatory Electrocardiogram (ECG)", Sensors, vol. 20, no. 5, pp.1461, 2020, doi: 10.3390/s20051461.
[5] M. Sraitih, Y. Jabrane, A. Hajjam El Hassani, "An automated system for ECG arrhythmia detection using machine learning techniques", Journal of Clinical Medicine, vol. 10, no. 22, pp. 5450, 2021, doi: 10.3390/jcm10225450.
[6] Z. Chen, H. Xu, J. Luo, T. Zhu, J. Meng, "Low-power perceptron model based ECG processor for premature ventricular contraction detection", Microprocessors and Microsystems, vol. 59, pp. 29-36, 2018, doi:10.1016/j.micpro.2018.03.006.
[7] A. Kumar, M. Singh ," Statistical analysis of ST segments in ECG signals for detection of ischaemic episodes",Transactions of the Institute of Measurement and Control, vol. 40, no. 3, pp. 819-830, 2018, doi: 10.1177/0142331216667811.
[8] S. H. Serhal, N. Abdallah, J. M. Marion, P. Chauvet, "Overview on prediction, detection, and classification of atrial fibrillation using wavelets and AI on ECG", Computers in Biology and Medicine, vol. 142, pp. 105168, 2022, doi: 10.1016/j.compbiomed.2021.105168.
[9] M. R. Rajeshwari and K. S. Kavitha, "Arrhythmia ventricular fibrillation classification on ECG signal using ensemble feature selection and deep neural network", Cluster Computing, vol. 25, no. 5, pp. 1-18, 2022, doi: 10.1007/s10586-022-03547-w.
[10] G. Jothi, J. Akilandeswari, S. David Samuel Azariya, A. Naveenkumar, "Unsupervised Feature Selection Approaches for Medical Dataset Using Soft Computing Techniques", in Evolution in Computational Intelligence,vol. 267, pp. 105-114, 2022, doi: 10.1007/978-981-16-6616-2_10.
[11] R. Allami, "Premature ventricular contraction analysis for real-time patient monitoring",Biomedical Signal Processing and Control,vol. 47, pp. 358-365, 2019, doi: 10.1016/j.bspc.2018.08.040.
[12] A. K. Bhoi, K. S. Sherpa, B. Khandelwal, "Arrhythmia and ischemia classification and clustering using QRS-ST-T (QT) analysis of electrocardiogram" Cluster Computing, vol. 21, no. 1, pp. 1033-1044, 2018, doi: 10.1007/s10586-017-0957-6.
[13] F. Zhou, L. Jin, J. Dong," Premature ventricular contraction detection combining deep neural networks and rules inference", Artificial intelligence in medicine,vol. 79, pp. 42-51, 2017, doi: 10.1016/j.artmed.2017.06.004.
[14] B. M. Mathunjwa, Y.T. Lin, C. H. Lin, M. F. Abbod, J. Shingshieh," ECG arrhythmia classification by using a recurrence plot and convolutional neural network", Biomedical Signal Processing and Control, vol. 64, pp. 102262, 2021, doi: 10.1016/j.bspc.2020.102262.
[15] J. Wang, "Automated detection of premature ventricular contraction based on the improved gated recurrent unit network", Computer Methods and Programs in Biomedicine,vol. 208, pp. 106284, 2021, doi: 10.1016/j.cmpb.2021.106284.
[16] G. B. Moody, R. G. Mark, "The impact of the MIT-BIH arrhythmia database". IEEE Engineering in Medicine and Biology Magazine, vol. 20, no. 3, pp. 45-50, 2001, doi: 10.1109/51.932724.
[17] A. I. Malakhov, S. I. Schookin, V. I. Ivancov, A. N. Tikhomirov," A combined algorithm for identification and differentiation of atrial flutter and atrial fibrillation based on ECG analysis",Biomedical Engineering, vol. 47, pp. 14-17, 2013, doi: 10.1007/s10527-013-9324-y.
[18] T. Ince, S. Kiranyaz, M." Gabbouj, A generic and robust system for automated patient-specific classification of ECG signals", IEEE Transactions on Biomedical Engineering, vol. 56, no. 5, pp. 1415-1426, 2009, doi: 10.1109/TBME.2009.2013934.
[19] A. Chen, Y. Zhang, M. Zhang, W. Liu, S. Chang, H. Wang, J. He, Q. Huang, "A real time QRS detection algorithm based on ET and PD controlled threshold strategy",Sensors, vol. 20, no. 14, pp. 4003, 2020, doi: 10.3390/s20144003.
[20] C. C. Lin, H. Y. Chang, Y. H. Huang, C. Y. Yeh, "A novel wavelet-based algorithm for detection of QRS complex",Applied Sciences, vol. 9, no. 10, pp. 2142, 2019, doi: 10.3390/app9102142.
[21] C.H. Lin" Frequency-domain features for ecg beat discrimination using grey relational analysis-based classifer" Computers & Mathematics with Applications, vol. 55, no. 4, pp. 680-690, 2008, doi: 10.1016/j.camwa.2007.04.035.
[22] V. Gupta, M. Mittal, V. Mittal, N. K. Saxena, "A critical review of feature extraction techniques for ECG signal analysis",Journal of The Institution of Engineers (India): Series B, vol. 102, no. 5, pp. 1049-1060, 2021, doi: 10.1007/s40031-021-00606-5.
[23] L. D. Avendano-Valencia, J. I. Godino-Llorente, M. Blanco-Velasco, G. Castellanos-Dominguez , "Feature extraction from parametric time–frequency representations for heart murmur detection". Annals of Biomedical Engineering, vol. 38, no. 8, pp. 2716-2732, 2010, doi: 10.1007/s10439-010-0077-4.
[24] S. Ghodake, S. Ghumbre, S. Deshmukh," Optimized cardiovascular disease detection and features extraction algorithms from ECG data", International Journal of Advanced Computer Science and Applications, vol. 11, no. 8, 2020, doi: 10.14569/IJACSA.2020.0110826.
[25] S. N. Yu, M. Y. Lee, "Wavelet-based multiscale sample entropy and chaotic features for congestive heart failure recognition using heart rate variability", Journal of Medical and Biological Engineering, vol. 35, no. 3, pp. 338-347, 2015, doi: 10.1007/s40846-015-0035-6.
[26] H. C. Nejad, O. Khayat, J. Razjouyan, "Chaotic feature extraction and neuro-fuzzy classifier for ECG signal characterization", Biomedical Engineering: Applications, Basis and Communications, vol. 26, no. 03, pp. 1450038, 2014, doi: 10.4015/S1016237214500380.
[27] D. J. Rezende, S. Mohamed, D. Wierstra," Stochastic backpropagation and approximate inference in deep generative models",in International conference on machine learning, vol. 32, pp. 1278-1276, 2014 , doi:10.48550/arXiv.1401.4082
[28] C. Xu, C. Sun, G. Jiang, X. Chen, Q. He, P. Xie," Two-level multi-domain feature extraction on sparse representation for motor imagery classification", Biomedical Signal Processing and Control, vol. 62, pp. 102160, 2020, doi: 10.1016/j.bspc.2020.102160.
[29] M. Ebrahimpoor, M. Taghizadeh, MH. Fatehi, O. Mahdiyar, J. Jamali, " Premature Ventricular Contractions Detection by Multi-Domain Feature Extraction and Auto-Encoder-based Feature Reduction", Circuits, Systems, and Signal Processing, vol. 43, no.5, pp. 3279–3296, 2024, doi: 10.1007/s00034-024-02613-5.
[30] M. H. Mazidi, M. Eshghi, M. R. Raoufy," Detection of premature ventricular contraction (PVC) using linear and nonlinear techniques: an experimental study", Cluster Computing, vol. 23, no. 2, pp. 759-774, 2020, doi: 10.1007/s10586-019-02953-x.
[31] J. Yu, X. Wang, X. Chen, J. Guo, "Automatic premature ventricular contraction detection using deep metric learning and KNN", Biosensors, vol. 11,no. 3, pp. 69, 2021, doi: 10.3390/bios11030069.
[32] X. Zhang, M. Shafiq, G. Zheng, J. Wan and Z. Sun, "Premature Ventricular Contractions’ Detection Based on Active Learning",Scientific Programming, vol. 2021, no. 1, pp. 5556011, 2021, doi: 10.1155/2021/5556011.
[33] N. T. Sarshar, M. Mirzaei, "Premature Ventricular Contraction Recognition Based on a Deep Learning Approach", Journal of Healthcare Engineering, vol. 2022, no. 1, pp. 1450723, 2022, doi: 10.1155/2022/1450723.