Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural Networks
Background Accurate detection of arrhythmic events in the intensive care units (ICU) is of paramount significance in providing timely care. However, traditional ICU monitors generate a high rate of false alarms causing alarm fatigue. In this work, we develop an algorithm to improve life threatening...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2021-12-01
|
| Series: | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| Subjects: | |
| Online Access: | https://www.ahajournals.org/doi/10.1161/JAHA.121.023222 |
| _version_ | 1797942685829431296 |
|---|---|
| author | Sandeep Chandra Bollepalli Rahul K. Sevakula Wan‐Tai M. Au‐Yeung Mohamad B. Kassab Faisal M. Merchant George Bazoukis Richard Boyer Eric M. Isselbacher Antonis A. Armoundas |
| author_facet | Sandeep Chandra Bollepalli Rahul K. Sevakula Wan‐Tai M. Au‐Yeung Mohamad B. Kassab Faisal M. Merchant George Bazoukis Richard Boyer Eric M. Isselbacher Antonis A. Armoundas |
| author_sort | Sandeep Chandra Bollepalli |
| collection | DOAJ |
| description | Background Accurate detection of arrhythmic events in the intensive care units (ICU) is of paramount significance in providing timely care. However, traditional ICU monitors generate a high rate of false alarms causing alarm fatigue. In this work, we develop an algorithm to improve life threatening arrhythmia detection in the ICUs using a deep learning approach. Methods and Results This study involves a total of 953 independent life‐threatening arrhythmia alarms generated from the ICU bedside monitors of 410 patients. Specifically, we used the ECG (4 channels), arterial blood pressure, and photoplethysmograph signals to accurately detect the onset and offset of various arrhythmias, without prior knowledge of the alarm type. We used a hybrid convolutional neural network based classifier that fuses traditional handcrafted features with features automatically learned using convolutional neural networks. Further, the proposed architecture remains flexible to be adapted to various arrhythmic conditions as well as multiple physiological signals. Our hybrid‐ convolutional neural network approach achieved superior performance compared with methods which only used convolutional neural network. We evaluated our algorithm using 5‐fold cross‐validation for 5 times and obtained an accuracy of 87.5%±0.5%, and a score of 81%±0.9%. Independent evaluation of our algorithm on the publicly available PhysioNet 2015 Challenge database resulted in overall classification accuracy and score of 93.9% and 84.3%, respectively, indicating its efficacy and generalizability. Conclusions Our method accurately detects multiple arrhythmic conditions. Suitable translation of our algorithm may significantly improve the quality of care in ICUs by reducing the burden of false alarms. |
| first_indexed | 2024-04-10T20:12:33Z |
| format | Article |
| id | doaj.art-91be6bb3a0174c2589222911f28ce759 |
| institution | Directory Open Access Journal |
| issn | 2047-9980 |
| language | English |
| last_indexed | 2024-04-10T20:12:33Z |
| publishDate | 2021-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| spelling | doaj.art-91be6bb3a0174c2589222911f28ce7592023-01-26T10:36:40ZengWileyJournal of the American Heart Association: Cardiovascular and Cerebrovascular Disease2047-99802021-12-01102310.1161/JAHA.121.023222Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural NetworksSandeep Chandra Bollepalli0Rahul K. Sevakula1Wan‐Tai M. Au‐Yeung2Mohamad B. Kassab3Faisal M. Merchant4George Bazoukis5Richard Boyer6Eric M. Isselbacher7Antonis A. Armoundas8Cardiovascular Research Center Massachusetts General Hospital Boston MACardiovascular Research Center Massachusetts General Hospital Boston MACardiovascular Research Center Massachusetts General Hospital Boston MACardiovascular Research Center Massachusetts General Hospital Boston MACardiology Division Emory University School of Medicine Atlanta GASecond Department of Cardiology Evangelismos General Hospital of Athens Athens GreeceAnesthesia Department Massachusetts General Hospital Boston MAHealthcare Transformation Lab Massachusetts General Hospital Boston MACardiovascular Research Center Massachusetts General Hospital Boston MABackground Accurate detection of arrhythmic events in the intensive care units (ICU) is of paramount significance in providing timely care. However, traditional ICU monitors generate a high rate of false alarms causing alarm fatigue. In this work, we develop an algorithm to improve life threatening arrhythmia detection in the ICUs using a deep learning approach. Methods and Results This study involves a total of 953 independent life‐threatening arrhythmia alarms generated from the ICU bedside monitors of 410 patients. Specifically, we used the ECG (4 channels), arterial blood pressure, and photoplethysmograph signals to accurately detect the onset and offset of various arrhythmias, without prior knowledge of the alarm type. We used a hybrid convolutional neural network based classifier that fuses traditional handcrafted features with features automatically learned using convolutional neural networks. Further, the proposed architecture remains flexible to be adapted to various arrhythmic conditions as well as multiple physiological signals. Our hybrid‐ convolutional neural network approach achieved superior performance compared with methods which only used convolutional neural network. We evaluated our algorithm using 5‐fold cross‐validation for 5 times and obtained an accuracy of 87.5%±0.5%, and a score of 81%±0.9%. Independent evaluation of our algorithm on the publicly available PhysioNet 2015 Challenge database resulted in overall classification accuracy and score of 93.9% and 84.3%, respectively, indicating its efficacy and generalizability. Conclusions Our method accurately detects multiple arrhythmic conditions. Suitable translation of our algorithm may significantly improve the quality of care in ICUs by reducing the burden of false alarms.https://www.ahajournals.org/doi/10.1161/JAHA.121.023222convolutional neural networksfalse alarmsintensive care unit monitorsmachine learningmulti‐class classification |
| spellingShingle | Sandeep Chandra Bollepalli Rahul K. Sevakula Wan‐Tai M. Au‐Yeung Mohamad B. Kassab Faisal M. Merchant George Bazoukis Richard Boyer Eric M. Isselbacher Antonis A. Armoundas Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural Networks Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease convolutional neural networks false alarms intensive care unit monitors machine learning multi‐class classification |
| title | Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural Networks |
| title_full | Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural Networks |
| title_fullStr | Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural Networks |
| title_full_unstemmed | Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural Networks |
| title_short | Real‐Time Arrhythmia Detection Using Hybrid Convolutional Neural Networks |
| title_sort | real time arrhythmia detection using hybrid convolutional neural networks |
| topic | convolutional neural networks false alarms intensive care unit monitors machine learning multi‐class classification |
| url | https://www.ahajournals.org/doi/10.1161/JAHA.121.023222 |
| work_keys_str_mv | AT sandeepchandrabollepalli realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT rahulksevakula realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT wantaimauyeung realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT mohamadbkassab realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT faisalmmerchant realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT georgebazoukis realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT richardboyer realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT ericmisselbacher realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks AT antonisaarmoundas realtimearrhythmiadetectionusinghybridconvolutionalneuralnetworks |