Premature ventricular contractions (PVCs) are abnormal heart beats that indicate potential heart
diseases. Diagnosis of PVCs is made by physicians examining long recordings of electrocardiogram (ECG), which
is onerous and time-consuming. In this study, deep learning was applied to develop models that can detect PVCs
in children automatically. This computer-aided diagnosis model achieved high accuracy while sustained stable
performance. It could save time and repeated efforts for physicians, enabling them to focus on more complicated
tasks.This study is a first step toward children’s PVC auto-detection in clinics. Further study will improve the
model’s performance with optimized structure and more data in different sources, while facing the challenges of
the variety and uncertainty of children’s ECG with heart diseases.
LIU Yixiua (刘宜修), HUANG Yujuanb (黄玉娟), WANG Jianyib (王健怡),LIU Lib (刘莉), LUO Jiajiaa* (罗家佳)
. Detecting Premature Ventricular Contraction in Children with Deep Learning[J]. Journal of Shanghai Jiaotong University(Science), 2018
, 23(1)
: 66
-73
.
DOI: 10.1007/s12204-018-1911-3
[1] EUGENIO P. Frequent premature ventricular contractions— An electrical link to cardiomyopathy [J]. Cardiologyin Review, 2015, 23: 168-172.
[2] BERTELS R A, HARTEVELD L M, FILIPPINI L H,et al. Left ventricular dysfunction is associated withfrequent premature ventricular complexes and asymptomaticventricular tachycardia in children [J]. EP Europace,2017, 19(4): 617-621.
[3] THANAPATAY D, SUWANSAROJ C, THANAWATTANDC. ECG beat classification method for ECGprintout with principle components analysis and supportvector machines [C]//2010 International Conferenceon Electronics and Information Engineering. Kyoto:IEEE, 2010: 72-75.
[4] KAYA Y, PHELIVAN H. Classification of prematureventricular contraction in ECG [J]. International Journalof Advanced Computer Science and Applications,2015, 6(7): 34-40.
[5] INAN O T, GIOVANGRANDI L, KOVACS G T A.Robust neural-network-based classification of prematureventricular contractions using wavelet transformand timing interval features [J]. IEEE Transactions onBiomedical Engineering, 2006, 53(12): 2507-2515.
[6] MAGLAVERAS N. ECG pattern recognition and classificationuing non-linear transformations and neuralnetworks: A review [J]. International Journal of MedicalInformatics, 1998, 52: 191-208.
[7] HUANHUAN M, YUE Z. Classification of electrocardiogramsignals with deep belief networks [C]//17thInternational Conference on Computational Scienceand Engineering, Chengdu: IEEE, 2014: 7-12.
[8] ZHOU F, JIN L, DONG J. Premature ventricular contractiondetection combining deep neural networks andrules inference [J]. Artificial Intelligence in Medicine,2017, 79: 42-51.
[9] CHRISTOV I, JEKOVA I, BORTOLAN G. Prematureventricular contraction classification by theKth nearest-neighbours rule [J]. Physiological Measurement,2005, 26: 123-130.
[10] RAHHALMMA, BAZI Y, ALHICHRIH, et al. Deeplearning approach for active classification of electrocardiogramsignals [J]. Information Sciences, 2016, 345:340-354.
[11] BORTOLAN G, JEKOVA I, CHRISTOV I. Comparisonof four methods for premature ventricular contractionand normal beat clustering [J]. Computers inCardiology, 2005, 32: 921-924.
[12] JIN L, DONG J. Ensemble deep learning for biomedicaltime series classification [J]. Computational Intelligenceand Neuroscience, 2016, 2016: 6212684.
[13] RAJPURKAR P, HANNUN A Y, HAGHPANAHIM, et al. Cardiologist-level arrhythmia detection withconvolutional neural networks [EB/OL]. (2017-07-06)[2017-09-25]. https://arxiv.org/pdf/1707.01836.pdf.
[14] SZEGEDY C, VANHOUCKE V, IOFFE S, et al. Rethinkingthe inception architecture for computer vision[C]//2016 IEEE Conference on Computer Visionand Pattern Recognition (CVPR). Las Vegas, Nevada:IEEE, 2016: 2818-2826.
