{"type":"ScholarlyArticle","title":"K-margin-based Residual-Convolution-Recurrent Neural Network for Atrial Fibrillation Detection","authors":[{"name":"Yuxi Zhou","slug":"yuxi-zhou","given":"Yuxi","family":"Zhou","affiliation":"School of Electronics Engineering and Computer Science , Peking University , Beijing , China; Key Laboratory of Machine Perception ( Peking University) , Ministry of Education , Beijing , China"},{"name":"Shenda Hong","slug":"shenda-hong","given":"Shenda","family":"Hong","affiliation":"School of Electronics Engineering and Computer Science , Peking University , Beijing , China; Key Laboratory of Machine Perception ( Peking University) , Ministry of Education , Beijing , China"},{"name":"Junyuan Shang","slug":"junyuan-shang","given":"Junyuan","family":"Shang","affiliation":"School of Electronics Engineering and Computer Science , Peking University , Beijing , China; Key Laboratory of Machine Perception ( Peking University) , Ministry of Education , Beijing , China"},{"name":"Meng Wu","slug":"meng-wu","given":"Meng","family":"Wu","affiliation":"School of Electronics Engineering and Computer Science , Peking University , Beijing , China; Key Laboratory of Machine Perception ( Peking University) , Ministry of Education , Beijing , China"},{"name":"Qingyun Wang","slug":"qingyun-wang","given":"Qingyun","family":"Wang","affiliation":"School of Electronics Engineering and Computer Science , Peking University , Beijing , China; Key Laboratory of Machine Perception ( Peking University) , Ministry of Education , Beijing , China"},{"name":"Hongyan Li","slug":"hongyan-li","given":"Hongyan","family":"Li","affiliation":"School of Electronics Engineering and Computer Science , Peking University , Beijing , China; Key Laboratory of Machine Perception ( Peking University) , Ministry of Education , Beijing , China"},{"name":"Junqing Xie","slug":"junqing-xie","given":"Junqing","family":"Xie","affiliation":"Medical Informatics Center , Peking University , Beijing , China"}],"abstract":"Atrial Fibrillation (AF) is an abnormal heart rhythm which can trigger cardiac arrest and sudden death. Nevertheless, its interpretation is mostly done by medical experts due to high error rates of computerized interpretation. One study found that only about 66% of AF were correctly recognized from noisy ECGs. This is in part due to insufficient training data, class skewness, as well as semantical ambiguities caused by noisy segments in an ECG record. In this paper, we propose a K-margin-based Residual-Convolution-Recurrent neural network (K-margin-based RCR-net) for AF detection from noisy ECGs. In detail, a skewness-driven dynamic augmentation method is employed to handle the problems of data inadequacy and class imbalance. A novel RCR-net is proposed to automatically extract both long-term rhythm-level and local heartbeat-level characters. Finally, we present a K-margin-based diagnosis model to automatically focus on the most important parts of an ECG record and handle noise by naturally exploiting expected consistency among the segments associated for each record. The experimental results demonstrate that the proposed method with 0.8125 F1NAOP score outperforms all state-of-the-art deep learning methods for AF detection task by 6.8%.","doi":"10.24963/ijcai.2019/839","url":"https://ijcai.org/proceedings/2019/839","pdf":"https://ijcai.org/proceedings/2019/0839.pdf","markdown":"https://ijcai.org/proceedings/2019/839.md","bibtex":"https://ijcai.org/proceedings/2019/bibtex/839","volume":{"slug":"2019","title":"Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2019","url":"https://ijcai.org/proceedings/2019/","year":2019,"location":"Macao","dates":"10-16 August 2019","editors":"Sarit Kraus","isbn":"978-0-9992411-4-1","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2019"},"section":"AI for Improving Human Well-being","track":"AI for Improving Human Well-being","number":839,"pages":{"first":6057,"last":6063},"keywords":["Special Track on AI for Improving Human-Well Being: Health applications (Special Track on AI and Human Wellbeing)"],"videos":[],"references":[{"position":1,"raw":"[Andreotti et al., 2017] Fernando Andreotti, Oliver Carr, and Marco AF Pimentel, et al. Comparing feature-based clas- sifiers and convolutional neural networks to detect arrhyth- mia from short segments of ecg. In CinC, pages 1-4. IEEE, 2017.","title":"Comparing feature-based classifiers and convolutional neural networks to detect arrhyth-mia from short segments of ecg","authors":"Fernando Andreotti, Oliver Carr, Marco Af Pimentel","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Carrara et al., 2015] Marta Carrara, Luca Carozzi, and Travis J Moss, et al. Heart rate dynamics distinguish among atrial fibrillation, normal sinus rhythm and sinus rhythm with frequent ectopy. Physiological measurement, 36(9):1873, 2015.","title":"Heart rate dynamics distinguish among atrial fibrillation, normal sinus rhythm and sinus rhythm with frequent ectopy","authors":"Marta Carrara, Luca Carozzi, Travis J Moss","year":"2015","venue":"Physiological measurement","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Chandra et al., 2017] BS Chandra, Challa S Sastry, and Soumya Jana, et al. Atrial fibrillation detection using con- volutional neural networks. In CinC, pages 1-4. IEEE, 2017.","title":"Atrial fibrillation detection using convolutional neural networks","authors":"Chandra","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Clifford et al., 2017] Gari D Clifford, Chengyu Liu, and Benjamin Moody, et al. Af classification from a short sin- gle lead ecg recording: the physionet/computing in cardi- ology challenge 2017. In CinC, pages 1-4. IEEE, 2017.","title":"Af classification from a short single lead ecg recording: the physionet/computing in cardiology challenge 2017","authors":"Chengyu Gari D Clifford, Benjamin Liu, Moody","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Gradl et al., 2012] Stefan Gradl, Patrick Kugler, and Clemens Lohmüller, et al. Real-time ecg monitoring and arrhythmia detection using android-based mobile devices. In EMBC, pages 2452-2455. IEEE, 2012.","title":"Real-time ecg monitoring and arrhythmia detection using android-based mobile devices","authors":"Stefan Gradl, Patrick Kugler, Clemens Lohmüller","year":"2012","venue":"EMBC","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Hannun et al., 2019] Awni Y Hannun, Pranav Rajpurkar, and Masoumeh Haghpanahi, et al. Cardiologist-level ar- rhythmia detection and classification in ambulatory elec- trocardiograms using a deep neural network. Nature medicine, 25(1):65, 2019.","title":"Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network","authors":"Pranav Awni Y Hannun, Masoumeh Rajpurkar, Haghpanahi","year":"2019","venue":"Nature medicine","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[He et al., 2016] Kaiming He, Xiangyu Zhang, and Shao- qing Ren, et al. Deep residual learning for image recogni- tion. In CVPR, pages 770-778. IEEE, 2016.","title":"Deep residual learning for image recognition","authors":"Kaiming He, Xiangyu Zhang, Shaoqing Ren","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Hong et al., 2017] Shenda Hong, Meng Wu, and Yuxi Zhou, et al. Encase: An ensemble classifier for ecg classification using expert features and deep neural networks. In CinC, pages 1-4. IEEE, 2017.","title":"Encase: An ensemble classifier for ecg classification using expert features and deep neural networks","authors":"Shenda Hong, Meng Wu, Yuxi Zhou","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Hong et al., 2019] Shenda Hong, Yuxi Zhou, and Meng Wu, et al. Combining deep neural networks and engineered fea- tures for cardiac arrhythmia detection from ECG record- ings. Physiological Measurement, 40(5):054009, 2019.","title":"Combining deep neural networks and engineered features for cardiac arrhythmia detection from ECG recordings","authors":"Shenda Hong, Yuxi Zhou, Meng Wu","year":"2019","venue":"Physiological Measurement","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Jiménez-Serrano et al., 2017] Santiago Jiménez-Serrano, Jaime Yagüe-Mayans, and Elena Simarro-Mondéjar, et al. Atrial fibrillation detection using feedforward neural networks and automatically extracted signal features. In CinC, pages 1-4. IEEE, 2017.","title":"Atrial fibrillation detection using feedforward neural networks and automatically extracted signal features","authors":"Jiménez-Serrano","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Kamaleswaran et al., 2018] Rishikesan Kamaleswaran, Ruhi Mahajan, and Oguz Akbilgic. A robust deep convolutional neural network for the classification of abnormal cardiac rhythm using single lead electrocardio- grams of variable length. Physiological measurement, 39(3):035006, 2018.","title":"Rishikesan Kamaleswaran, Ruhi Mahajan, and Oguz Akbilgic. A robust deep convolutional neural network for the classification of abnormal cardiac rhythm using single lead electrocardiograms of variable length","authors":"Kamaleswaran","year":"2018","venue":"Physiological measurement","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Kass and Clancy, 2005] Robert E Kass and Colleen E Clancy. Basis and treatment of cardiac arrhythmias, vol- ume 171. Springer Science & Business Media, 2005.","title":"Basis and treatment of cardiac arrhythmias","authors":"E Robert, Colleen E Kass, Clancy","year":"2005","venue":null,"doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Kiranyaz et al., 2015] Serkan Kiranyaz, Turker Ince, and Moncef Gabbouj. Real-time patient-specific ecg classifi- cation by 1-d convolutional neural networks. IEEE Trans- actions on Biomedical Engineering, 63(3):664-675, 2015.","title":"Real-time patient-specific ecg classification by 1-d convolutional neural networks","authors":"Serkan Kiranyaz, Turker Ince, Moncef Gabbouj","year":"2015","venue":"IEEE Transactions on Biomedical Engineering","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Linker, 2016] David Linker. Accurate, automated detec- tion of atrial fibrillation in ambulatory recordings. CVET, 7(2):182-189, 2016.","title":"Accurate, automated detection of atrial fibrillation in ambulatory recordings","authors":"David Linker, Linker","year":"2016","venue":"CVET","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Majumdar and Ward, 2017] Angshul Majumdar and Rabab Ward. Robust greedy deep dictionary learning for ecg arrhythmia classification. In IJCNN, pages 4400-4407. IEEE, 2017.","title":"Robust greedy deep dictionary learning for ecg arrhythmia classification","authors":"Ward ; Majumdar, Angshul Majumdar, Rabab Ward","year":"2017","venue":"IJCNN","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Schwab et al., 2017] Patrick Schwab, Gaetano C Scebba, and Jia Zhang, et al. Beat by beat: Classifying cardiac ar- rhythmias with recurrent neural networks. In CinC, pages 1-4. IEEE, 2017.","title":"Beat by beat: Classifying cardiac arrhythmias with recurrent neural networks","authors":"Patrick Schwab, C Gaetano, Jia Scebba, Zhang","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Shah and Rubin, 2007] Atman P Shah and Stanley A Ru- bin. Errors in the computerized electrocardiogram inter- pretation of cardiac rhythm. Journal of electrocardiology, 40(5):385-390, 2007.","title":"Errors in the computerized electrocardiogram interpretation of cardiac rhythm","authors":"P Atman, Stanley A Shah, Rubin","year":"2007","venue":"Journal of electrocardiology","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Silva et al., 2011] Ikaro Silva, George B Moody, and Leo Celi. Improving the quality of ecgs collected using mobile phones: The physionet/computing in cardiology challenge 2011. In CinC, pages 273-276. IEEE, 2011.","title":"Improving the quality of ecgs collected using mobile phones: The physionet/computing in cardiology challenge","authors":"Ikaro Silva, George B Moody, Leo Celi","year":"2011","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Sodmann et al., 2018] Philipp Sodmann, Marcus Vollmer, and Neetika Nath, et al. A convolutional neural network for ecg annotation as the basis for classification of cardiac rhythms. Physiological measurement, 39(10):104005, 2018.","title":"A convolutional neural network for ecg annotation as the basis for classification of cardiac rhythms","authors":"Philipp Sodmann, Marcus Vollmer, Neetika Nath","year":"2018","venue":"Physiological measurement","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Stepien and Grzegorczyk, 2017] Katarzyna Stepien and Iga Grzegorczyk. Classification of ecg recordings with neu- ral networks based on specific morphological features and regularity of the signal. In CinC, pages 1-4. IEEE, 2017.","title":"Classification of ecg recordings with neural networks based on specific morphological features and regularity of the signal","authors":"Katarzyna Stepien, Iga Grzegorczyk","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Tateno and Glass, 2001] K Tateno and L Glass. Automatic detection of atrial fibrillation using the coefficient of varia- tion and density histograms of rr and δrr intervals. MBEC, 39(6):664-671, 2001.","title":"Automatic detection of atrial fibrillation using the coefficient of variation and density histograms of rr and δrr intervals","authors":"K Tateno, L Glass","year":"2001","venue":"MBEC","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Warrick and Homsi, 2018] Philip A Warrick and Ma- sun Nabhan Homsi. Ensembling convolutional and long short-term memory networks for electrocardio- gram arrhythmia detection. Physiological measurement, 39(11):114002, 2018.","title":"Ensembling convolutional and long short-term memory networks for electrocardiogram arrhythmia detection","authors":"Homsi ; Warrick, Sun Nabhan, Homsi","year":"2018","venue":"Physiological measurement","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Wu et al., 2016] Zhiyong Wu, Xiangqian Ding, and Guan- grui Zhang. A novel method for classification of ecg arrhythmias using deep belief networks. IJCIA, 15(04):1650021, 2016.","title":"A novel method for classification of ecg arrhythmias using deep belief networks","authors":"Zhiyong Wu, Xiangqian Ding, Guangrui Zhang","year":"2016","venue":"IJCIA","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Xiong et al., 2017] Zhaohan Xiong, Martin K Stiles, and Jichao Zhao. Robust ecg signal classification for detec- tion of atrial fibrillation using a novel neural network. In CinC, pages 1-4. IEEE, 2017.","title":"Robust ecg signal classification for detection of atrial fibrillation using a novel neural network","authors":"Xiong","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Ye et al., 2010] Can Ye, Miguel Tavares Coimbra, and BVK Vijaya Kumar. Arrhythmia detection and classifi- cation using morphological and dynamic features of ecg signals. In EMBC, pages 1918-1921. IEEE, Aug 2010.","title":"Arrhythmia detection and classification using morphological and dynamic features of ecg signals","authors":"Ye","year":"2010","venue":"EMBC","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Zihlmann et al., 2017] Martin Zihlmann, Dmytro Perekrestenko, and Michael Tschannen. Convolu- tional recurrent neural networks for electrocardiogram classification. In CinC, pages 1-4. IEEE, 2017.","title":"Convolutional recurrent neural networks for electrocardiogram classification","authors":"Martin Zihlmann, Dmytro Perekrestenko, Michael Tschannen","year":"2017","venue":"CinC","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}