本研究提出了結合更快速區域卷積類神經網路與深度卷積類神經網路的異常心電圖偵測演算法，利用更快速區域卷積類神經網路在物件辨識上高度準確率特性做QRS複合波的定位，而後再利用卷積類神經網路在影像資料上優秀的分類準確率做異常心電圖的偵測。本研究使用MIT-BIH心律不整資料庫中的Lead-2的心電圖資料作為資料來源，利用五秒的訊號圖做為QRS複合波定位的資料集，在十倍交叉驗證下，QRS複合波的定位準確率可達99.5%。在異常心電圖偵測中，本研究除了利用原訊號圖外，也利用連續小波轉換與主成分分析作特徵提取，以提升分類準確率，在分類結果中，總共有三種分類結果，分別是三類疾病(SR, VPC, APC)分類結果、五類疾病(SR, APC, VPC, LBBB, RBBB)分類結果、八類疾病(SR, VPC, APC, LBBB, RBBB, paced, AAPC, FVN)分類結果，三種分類結果的最高準確率分別為99.92%、99.31%、98.78%。

This study proposes an abnormal ECG detection algorithm that combines a faster regional convolutional neural network and a deep convolutional neural network and uses the characteristics of high accuracy in object detection using faster regional convolutional neural networks to do QRS complex detection. Then use the convolutional neural network to classify the image data for the detection of abnormal ECG. In this study, the Lead-2 ECG data in the MIT-BIH arrhythmia database was used as training data, and the five-second signal figure was used as the training data set for QRS complex detection. Under ten-fold cross-validation, the average accuracy can reach 99.5%. In the detection of abnormal ECG, in addition to the original signal map, this study also uses continuous wavelet transform and principal component analysis for feature extraction to improve the classification accuracy. Among the classification results, there are three classification results, which are three types of diseases (SR, VPC, APC) classification results, five types of diseases (SR, APC, VPC, LBBB, RBBB) classification results, eight types of diseases (SR, VPC, APC, LBBB, RBBB, paced, AAPC, FVN) classification results, the highest average accuracy of the three classification results are 99.92%, 99.31%, 98.78%.

[1] Arbateni Khaled and Abdelhak Bennia, “Sigmoidal radial basis function ANN for QRS complex detection,” Neurocomputing, vol. 145, pp. 438-450, 2014.
[2] Saini Indu, Dilbag Singh, and Arun Khosla, “QRS detection using K-Nearest Neighbor algorithm (KNN) and evaluation on standard ECG databases,” Journal of advanced research, vol. 4, no. 4, pp. 331-344, 2013.
[3] Sarabjeet S. Mehta, Dipali Shete, Nitin S. Lingayat, and V.S. Chouhan, “K-means algorithm for the detection and delineation of QRS-complexes in Electrocardiogram,” Irbm, vol. 31, no. 1, pp. 48-54, 2010.
[4] Jalil Bushra, Laligant Olivier, Fauvet Eric, and Beya Ouadi, “Detection of QRS complex in ECG signal based on classification approach,” 2010 IEEE International Conference on Image Processing, pp. 345-348, 2010.
[5] Xiang Yande, Zhitao Lin, and Jianyi Meng, “Automatic QRS complex detection using two-level convolutional neural network,” Biomedical engineering online, vol. 17, no. 13, 2018.
[6] Yu Runze, Gao Yingguo, Duan Xiaohui, Zhu Tiangang, Wang Zhilong, and Jiao Bingli, “QRS detection and measurement method of ECG paper based on convolutional neural networks,” 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 4636-4639, 2018.
[7] Cai Wenjie and Danqin Hu., “QRS complex detection using novel deep learning neural networks,” IEEE Access, 2020.
[8] Silva Pedro, Luz Eduardo, Wanner Elizabeth, Menotti David, and Moreira Gladston, “QRS Detection in ECG Signal with Convolutional Network,” Iberoamerican Congress on Pattern Recognition, pp. 802-809, 2018.
[9] Habib Ahsan, Chandan Karmakar, and John Yearwood, “Impact of ECG dataset diversity on generalization of CNN model for detecting QRS complex,” IEEE access,
vol. 7, pp. 93275-93285, 2019.
[10] Yang Weiyi, Si Yujuan, Wang Di, and Guo Buhao, “Automatic recognition of arrhythmia based on principal component analysis network and linear support vector machine,” Computers in biology and medicine, vol. 101, pp. 22-32, 2018.
[11] Khalaf Aya F., Mohamed I. Owis, and Inas A. Yassine, “A novel technique for cardiac arrhythmia classification using spectral correlation and support vector machines,” Expert Systems with Applications, vol. 42, no. 21, pp. 8361-8368, 2015.
[12] Elhaj Fatin A., Salim Naomie, Harris Arief R., Swee Tan Tian, and Ahmed Taqwa “Arrhythmia recognition and classification using combined linear and nonlinear features of ECG signals,” Computer methods and programs in biomedicine, vol. 127, pp. 52-63, 2016.
[13] Martis Roshan Joy, U. Rajendra Acharya, Lim Choo Min, Mandana K. M., Ray A. K., and Chakraborty Chandan, “Application of higher-order cumulant features for cardiac health diagnosis using ECG signals,” International journal of neural systems, vol. 23, no. 04, pp. 1350014.
[14] Güler İnan and Elif Derya Übeylı, “ECG beat classifier designed by combined neural network model,” Pattern recognition, vol. 38, no. 2, pp. 199-208, 2015.
[15] Martis, Roshan Joy, U. Rajendra Acharya, and Lim Choo Min, “ECG beat classification using PCA, LDA, ICA, and discrete wavelet transform,” Biomedical Signal Processing and Control, vol. 8, no. 5, pp. 437-448, 2013.
[16] Shadmand Shirin and Behbood Mashoufi, “A new personalized ECG signal classification algorithm using block-based neural network and particle swarm optimization,” Biomedical Signal Processing and Control, vol. 25, pp. 12-23, 2016.
[17] Sahoo Santanu, Kanungo Bhupen, Behera Suresh, and Sabut Sukanta, “Multiresolution wavelet transform based feature extraction and ECG classification to detect cardiac abnormalities,” Measurement, vol. 108, pp. 55-66, 2017.
[18] Afkhami Rashid Ghorbani, Ghanbar Azarnia, and Mohammad Ali Tinati, “Cardiac arrhythmia classification using statistical and mixture modeling features of ECG signals,” Pattern Recognition Letters, vol. 70, pp. 45-51. 2016.
[19] Yildirim Özal, “A novel wavelet sequence based on deep bidirectional LSTM network model for ECG signal classification,” Computers in biology and medicine, vol. 96, pp. 189-202, 2018.
[20] Oh Shu Lih, Ng Eddie Y.K., Tan Ru San, and Acharya U. Rajendra, “Automated diagnosis of arrhythmia using combination of CNN and LSTM techniques with variable length heartbeats,” Computers in biology and medicine, vol. 102, pp. 278-287, 2018.
[21] Yıldırım Özal, Pławiak Paweł, Tan Ru-San, and Acharya U. Rajendra, “Arrhythmia detection using deep convolutional neural network with long-duration ECG signals,” Computers in biology and medicine, vol. 102,
pp. 411-420, 2018.
[22] Faust Oliver, Shenfield Alex, Kareem Murtadha, San Tan Ru, Fujita Hamido, and Acharya U. Rajendra, “Automated detection of atrial fibrillation using long short-term memory network with RR interval signals,” Computers in biology and medicine, vol. 102, pp. 327-335, 2018.
[23] Hannun Awni Y., Rajpurkar Pranav, Haghpanahi Masoumeh, Tison Geoffrey H., Bourn Codie, Turakhia Mintu P., and Ng Andrew Y., “Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network,” Nature medicine, vol. 25, no. 1, pp. 65, 2019.
[24] Huang Jingshan, Chen Binqiang, Yao Bin, and He Wangpeng, “ECG arrhythmia classification using STFT-based spectrogram and convolutional neural network,” IEEE Access, vol. 7, pp. 92871-92880, 2019.
[25] Oh Shu Lih, Ng Eddie Y.K., Tan Ru San, and Acharya U. Rajendra, “Automated beat-wise arrhythmia diagnosis using modified U-net on extended electrocardiographic recordings with heterogeneous arrhythmia types,” Computers in biology and medicine, vol. 105, pp. 92-101, 2019.
[26] Luo Kan, Li Jianqing, Wang Zhigang, and Cuschier Alfred, “Patient-specific deep architectural model for ECG classification,” Journal of healthcare engineering, 2017.
[27] Acharya U. Rajendra, Oh Shu Lih, Hagiwara Yuki, Tan Jen Hong, Adam Muhammad, and Gertych Arkadiusz, Tan Ru San, “A deep convolutional neural network model to classify heartbeats,” Computers in biology and medicine, vol. 89, pp. 389-396, 2017.
[28] Fujita, Hamido, and Dalibor Cimr, “Computer aided detection for fibrillations and flutters using deep convolutional neural network,” Information Sciences, vol. 486, pp. 231-239, 2019.
[29] Pławiak Paweł, “Novel genetic ensembles of classifiers applied to myocardium dysfunction recognition based on ECG signals,” Swarm and evolutionary computation,
vol. 39, pp. 192-208, 2018.
[30] Kiranyaz Serkan, Ince Turker, Pulkkinen Jenni, and Gabbouj Moncef, “Personalized long-term ECG classification: A systematic approach,” Expert Systems with Applications, vol. 38, no. 4, pp. 3220-3226, 2011.
[31] Ren Shaoqing, He Kaiming, Girshick Ross, and Sun Jian, “Faster r-cnn: Towards real-time object detection with region proposal networks,” Advances in neural information processing systems, pp. 91-99, 2015.
[32] Krizhevsky Alex, Ilya Sutskever, and Geoffrey E. Hinton, “Imagenet classification with deep convolutional neural networks,” Advances in neural information processing systems, pp. 1097-1105, 2012.
[33] Szegedy Christian, Liu Wei, Jia Yangqing, Sermanet Pierre, Reed Scott, Anguelov Dragomir, Erhan Dumitru, Vanhoucke Vincent, and Rabinovich Andrew, “Going deeper with convolutions,” Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1-9, 2015.
[34] He Kaiming, Zhang Xiangyu, Ren Shaoqing, and Sun Jian, “Deep residual learning for image recognition,” Proceedings of the IEEE conference on computer vision and pattern recognition, pp.770-778, 2016.
[35] “Continuous wavelet transform,” Wikipedia.com, April. 14, 2020. [online].
Available: https://en.wikipedia.org/wiki/Continuous_wavelet_transform. [Accessed: June. 4, 2020].
[36] Oh Hyun Min, Hyunki Lee, and Min Young Kim, "Comparing Convolutional Neural Network (CNN) models for machine learning-based drone and bird classification of anti-drone system." 2019 19th International Conference on Control, Automation and Systems (ICCAS), 2019.