Open AccessStudy on abnormal detection of ecg signal base on DCNN
Author(s) -
Kun Guo,
Lifeng Ma,
Qin JiaTong,
He Yunxia
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1345/4/042006
Subject(s) - qrs complex , convolutional neural network , computer science , artificial intelligence , pattern recognition (psychology) , dropout (neural networks) , test set , signal (programming language) , data set , base (topology) , set (abstract data type) , anomaly detection , deep learning , algorithm , mathematics , machine learning , cardiology , medicine , mathematical analysis , programming language
In this paper, the Dropout deep convolutional neural network based on ecg signal anomaly detection method is proposed. Then, the double threshold method is used to detect the R wave and calculate the width of corresponding QRS wave group and the position of QRS wave is taken as the reference to include several points forward and backward respectively. These data points are intercepted and made into data sets for training and testing of deep convolutional neural network. Finally, the data set is read into deep convolutional neural network for abnormal measurement, and the test results are statistically analysed. In this paper, the detection rate of QRS composite wave is up to 99.5%, and the classification accuracy of abnormal signals can reach 99.26%. Experimental results show that the algorithm has the characteristics of high implementation rate, high accuracy, convenient and fast.