ECG feature extraction via wavelet transform and machine learning algorithms
Yıl 2018,
Cilt: 20 Sayı: 1, 94 - 109, 09.04.2018
Hülya Kodal Sevindir
,
Süleyman Çetinkaya
,
Cüneyt Yazıcı
Öz
Nowadays usage of wavelet transform on biomedical signals has been increased and effective results have been obtained. In this study wavelet transform is used to obtain better results on QRS complex detection; wavelets are used to get rid of baseline wandering and high frequency artifact on ECG. To do so, some ECG recordings on MIT-BIH arrhythmia database are used. To omit baseline wandering Daubechies wavelet up to level 10 is used and approximation and detail coefficients at level 10 are excluded from the ECG signal. To omit high frequency artifact wavelet denoising is applied to the ECG signals. Afterward, using the first and second derivative information of the signal, Support Vector Machine and Naive Bayes algorithms are applied separately. According to the study conducted, although SVM algorithm runs slower than Naive Bayes, results for SVM are much better for QRS detection; the results for SVM are %99.46 sensitivity, %100 positive sensitivity, and %0.54 error.
Kaynakça
- https://www.medikalakademi.com.tr/kalp-hastaliklari-neden-genclerde-daha-olumcul/, (20.06.2017).
- Kodal Sevindir, H., Cetinkaya, S. ve Sayli, O., Wavelet transform based noise removal from ECG signal for accurate heart rate detection using ECG, Medical Technologies National Conference (TIPTEKNO), Muğla, (2015)
- Yanık, H. ve Değirmenci, E., Detection of ECG characteristic points using multiresolution analysis, Sinyal işleme ve iletişim uygulamaları (SİU), Malatya, 383-386, (2015).
- Jiang, X. ve Zhang, L., ECG arrhythmias recognition system based on independent component analysis feature extraction, IEEE Region 10 Conference, Hong Kong, (2006).
- http://tinaztepehastanesi.com.tr/saglik_kosesi/kardiyoloji/ekg-nedir, (18 Mart 2015).
- http://www.slideshare.net/husam685/ekg-ritim-ve-pace-maker, (25 Kasım 2016)
- Uslu, E. ve Bilgin, G., Dalgacık ve birleşik dalgacık paket dönüşümü kullanarak kalp aritmilerinin sınıflandırılması, IEEE 16. Sinyal İşleme ve Uygulamaları Kurultayı, Antalya, (2008).
- Smith, M.J., ve Barnwell, T.P., A procedure for designing exact reconstruction filter banks for tree structured sub-band coders, In Proc. IEEE Int.Conf. Acoust., Speech, and Signal Proc., San Diego, (1984).
- Turan, M.D., EKG Sinyalindeki gürültülerin IIR filtreler ile matlabda filtrelenmesi, Bitirme Ödevi, Süleyman Demirel Üniversitesi, (2005).
- Vapnik, V., Statistical learning theory, Wiley Press, New York, (1998).
- Domingos, P. ve Pazzani, M., On the optimality of the simple Bayesian classifier under zero-one loss, Machine Learning, 29, 103–130, (1997).
- Goldberger, A.L., Amaral, L. A., Glass, L., Hausdorff, J. M., Ivanov, P. C, Mark, R. G., Mietus, J.E., Moody, G. B., Peng, C. K. ve Stanley, H. E., PhysioBank, PhysioToolkit, PhysioNet, components of a new research resource for complex physiologic signals, Circulation, 101, 23, 215–220, (2000).
- Sahambi, J.S., Tandon S.N. ve Bhatt, R. K. P., Using wavelet transforms for ECG characterization, IEEE Engineering in Medicine and Biology, 97, 77–83, (1997).
- Mehta, S.S. ve Lingayat, N.S., SVM-based algorithm for recognition of QRS complexes in electrocardiogram, Elsevier Masson IRBM, 29, 310–317, (2008).
- Sasikala, P. ve Wahidabanu, R.S.D., Robust QRS Peak and QRS detection in Electrocardiogram using Wavelet Transform, International Journal of Advanced Computer Science and Applications, 1, 6, 48–53, (2010).
- Dinh, H.A.N., Kumar, D.K., Pah, N.D. ve Burton, P., Wavelets for QRS Detection, Proceedings of the 23rd Annual EMBS International Conference, 1, 1883–1887, İstanbul, (2001).
- Xia, Y., Han, J. ve Wang, K., Quick detection of QRS complexes and R-waves using a wavelet transform and K-means clustering, Bio-Medical Materials and Engineering, 26, 1059–1065, (2015).
- Gritzali, F., Towards a Generalized Scheme For QRS Detection in ECG Waveforms, Signal Processing, 15, 183–192, (1988).
- Mehta, S.S., Shete, D.A., Lingayat, N.S. ve Chouhan, V.S., K-means algorithm for the detection and delineation of QRS-complexes in electrocardiogram, Elsevier Masson IRBM, 31, 48–54, (2010).
- Chouhan, V.S. ve Mehta, S.S., Detection of QRS complex in 12-lead ECG using adaptive quantized threshold, International Journal of Computer Science and Network Security, 8, 155–163, (2008).
- Mehta, S.S. ve Lingayat, N.S., Combined entropy based method for detection of QRS complexes in 12-lead electrocardiogram using SVM, Computers in Biology and Medicine, 38, 138–145, (2008).
- Gayake, M.A. ve Shete, V.V., ECG QRS-Complex Detection using SVM, ITSI Transactions on Electrical and Electronics Engineering, 2, 5-8, (2014).
- Singh, D. ve Khosla, A., QRS detection using K-Nearest Neighbor algorithm (KNN) and evaluation on standard ECG databases, Journal of Advanced Research, 4, 331–344, (2013).
Makine öğrenmesi algoritmaları ve dalgacık dönüşümü ile EKG sinyalinden özellik çıkarımı
Yıl 2018,
Cilt: 20 Sayı: 1, 94 - 109, 09.04.2018
Hülya Kodal Sevindir
,
Süleyman Çetinkaya
,
Cüneyt Yazıcı
Öz
Günümüzde biyomedikal sinyallerin analizinde dalgacık dönüşümünün
kullanılması oldukça yaygın olup elde edilen sonuçlar etkileyicidir. Bu
çalışmada, biyomedikal sinyallerden elektrokardiyogram (EKG) sinyallerinde QRS
zirvesi belirleme hedeflenmiş ve daha iyi sonuçlar almak için öncelikle EKG
sinyallerindeki zemin gezinme gürültüsünün giderilmesi ve yüksek frekanslı
gürültünün temizlenmesi amacıyla dalgacık analizi kullanılmıştır. Daubechies 10
(db10) dalgacık dönüşümü uygulanan sinyalin 10. seviye yaklaşım katsayısı ve
10. seviye detay katsayısı çıkartılarak sinyaldeki zemin gezinmesi problemi
giderilmistir. Yüksek frekans gürültüsünün giderilmesi için ise zemin gezinmesi
problemi giderilmiş olan sinyale dalgacık gürültü temizleme uygulanmıştır.
Gürültüsü temizlenen sinyalde QRS zirvelerini belirlemek için sinyalin 1. türev
ve 2. türev bilgileri ele alınarak Destek Vektör Makineleri ve Naive Bayes
algoritmaları kullanılmıştır. QRS zirvelerinin bulunmasında, MIT-BIH aritmi
veri tabanında verilen QRS zirvelerinin konum bilgileri kullanılmıştır. QRS
zirvelerini doğru belirlemede Destek Vektör Makineleri algoritması Naive Bayes
algoritmasından daha yavaş sonuç vermesine rağmen %99.46 duyarlılık, %100
seçicilik ve %0.54 hata değerlerine ulaşmıştır.
Kaynakça
- https://www.medikalakademi.com.tr/kalp-hastaliklari-neden-genclerde-daha-olumcul/, (20.06.2017).
- Kodal Sevindir, H., Cetinkaya, S. ve Sayli, O., Wavelet transform based noise removal from ECG signal for accurate heart rate detection using ECG, Medical Technologies National Conference (TIPTEKNO), Muğla, (2015)
- Yanık, H. ve Değirmenci, E., Detection of ECG characteristic points using multiresolution analysis, Sinyal işleme ve iletişim uygulamaları (SİU), Malatya, 383-386, (2015).
- Jiang, X. ve Zhang, L., ECG arrhythmias recognition system based on independent component analysis feature extraction, IEEE Region 10 Conference, Hong Kong, (2006).
- http://tinaztepehastanesi.com.tr/saglik_kosesi/kardiyoloji/ekg-nedir, (18 Mart 2015).
- http://www.slideshare.net/husam685/ekg-ritim-ve-pace-maker, (25 Kasım 2016)
- Uslu, E. ve Bilgin, G., Dalgacık ve birleşik dalgacık paket dönüşümü kullanarak kalp aritmilerinin sınıflandırılması, IEEE 16. Sinyal İşleme ve Uygulamaları Kurultayı, Antalya, (2008).
- Smith, M.J., ve Barnwell, T.P., A procedure for designing exact reconstruction filter banks for tree structured sub-band coders, In Proc. IEEE Int.Conf. Acoust., Speech, and Signal Proc., San Diego, (1984).
- Turan, M.D., EKG Sinyalindeki gürültülerin IIR filtreler ile matlabda filtrelenmesi, Bitirme Ödevi, Süleyman Demirel Üniversitesi, (2005).
- Vapnik, V., Statistical learning theory, Wiley Press, New York, (1998).
- Domingos, P. ve Pazzani, M., On the optimality of the simple Bayesian classifier under zero-one loss, Machine Learning, 29, 103–130, (1997).
- Goldberger, A.L., Amaral, L. A., Glass, L., Hausdorff, J. M., Ivanov, P. C, Mark, R. G., Mietus, J.E., Moody, G. B., Peng, C. K. ve Stanley, H. E., PhysioBank, PhysioToolkit, PhysioNet, components of a new research resource for complex physiologic signals, Circulation, 101, 23, 215–220, (2000).
- Sahambi, J.S., Tandon S.N. ve Bhatt, R. K. P., Using wavelet transforms for ECG characterization, IEEE Engineering in Medicine and Biology, 97, 77–83, (1997).
- Mehta, S.S. ve Lingayat, N.S., SVM-based algorithm for recognition of QRS complexes in electrocardiogram, Elsevier Masson IRBM, 29, 310–317, (2008).
- Sasikala, P. ve Wahidabanu, R.S.D., Robust QRS Peak and QRS detection in Electrocardiogram using Wavelet Transform, International Journal of Advanced Computer Science and Applications, 1, 6, 48–53, (2010).
- Dinh, H.A.N., Kumar, D.K., Pah, N.D. ve Burton, P., Wavelets for QRS Detection, Proceedings of the 23rd Annual EMBS International Conference, 1, 1883–1887, İstanbul, (2001).
- Xia, Y., Han, J. ve Wang, K., Quick detection of QRS complexes and R-waves using a wavelet transform and K-means clustering, Bio-Medical Materials and Engineering, 26, 1059–1065, (2015).
- Gritzali, F., Towards a Generalized Scheme For QRS Detection in ECG Waveforms, Signal Processing, 15, 183–192, (1988).
- Mehta, S.S., Shete, D.A., Lingayat, N.S. ve Chouhan, V.S., K-means algorithm for the detection and delineation of QRS-complexes in electrocardiogram, Elsevier Masson IRBM, 31, 48–54, (2010).
- Chouhan, V.S. ve Mehta, S.S., Detection of QRS complex in 12-lead ECG using adaptive quantized threshold, International Journal of Computer Science and Network Security, 8, 155–163, (2008).
- Mehta, S.S. ve Lingayat, N.S., Combined entropy based method for detection of QRS complexes in 12-lead electrocardiogram using SVM, Computers in Biology and Medicine, 38, 138–145, (2008).
- Gayake, M.A. ve Shete, V.V., ECG QRS-Complex Detection using SVM, ITSI Transactions on Electrical and Electronics Engineering, 2, 5-8, (2014).
- Singh, D. ve Khosla, A., QRS detection using K-Nearest Neighbor algorithm (KNN) and evaluation on standard ECG databases, Journal of Advanced Research, 4, 331–344, (2013).