Electrocardiogram(ECG)is widely used to detect arrhythmia.Atrial fibrillation,atrioventricular block,premature beats,etc.can all be diagnosed by ECG.When the distribution of training data and test data is inconsistent...Electrocardiogram(ECG)is widely used to detect arrhythmia.Atrial fibrillation,atrioventricular block,premature beats,etc.can all be diagnosed by ECG.When the distribution of training data and test data is inconsistent,the accuracy of the model will be affected.This phenomenon is called dataset shift.In the real-world heartbeat classification system,the heartbeat of the training set and test set often comes from patients of different ages and genders,so there are differences in the distribution of data sets.The main challenge in applying machine learning algorithms to clinical AI systems is dataset shift.Test-time adaptation(TTA)aims to adapt a pre-trained model from the source domain(SD)to the target domain(TD)without using any SD data or TD labels,thereby reducing model performance degradation due to domain differences.We propose a method based on multimodal image fusion and continual test-time adaptation(FCTA)for accurate and efficient heartbeat classification.First,the original ECG data is converted into a three-channel color image through a multimodal image fusion framework.The impact of class imbalance on network performance is overcome using a batch weight loss function,and then the pretrained source model is adapted to the TD using a continual test-time adaptation(CTA)method.Although our method is very simple,compared with other domain adaptation methods,it can significantly improve model performance on the test set and reduce the impact caused by the difference in domain distribution.展开更多
Arrhythmias may lead to sudden cardiac death if not detected and treated in time.A supraventricular premature beat(SPB)and premature ventricular contraction(PVC)are important categories of arrhythmia disease.Recently,...Arrhythmias may lead to sudden cardiac death if not detected and treated in time.A supraventricular premature beat(SPB)and premature ventricular contraction(PVC)are important categories of arrhythmia disease.Recently,deep learning methods have been applied to the PVC/SPB heartbeats detection.However,most researchers have focused on time-domain information of the electrocardiogram and there has been a lack of exploration of the interpretability of the model.In this study,we design an interpretable and accurate PVC/SPB recognition algorithm,called the interpretable multilevel wavelet decomposition deep network(IMWDDN).Wavelet decomposition is introduced into the deep network and the squeeze and excitation(SE)-Residual block is designed for extracting time-domain and frequency-domain features.Additionally,inspired by the idea of residual learning,we construct a novel loss function for the constant updating of the multilevel wavelet decomposition parameters.Finally,the IMWDDN is evaluated on the Third China Physiological Signal Challenge Dataset and the MIT-BIH Arrhythmia database.The comparison results show IMWDDN has better detection performance with 98.51%accuracy and a 93.75%F1-macro on average,and its areas of concern are similar to those of an expert diagnosis to a certain extent.Generally,the IMWDDN has good application value in the clinical screening of PVC/SPB heartbeats.展开更多
文摘Electrocardiogram(ECG)is widely used to detect arrhythmia.Atrial fibrillation,atrioventricular block,premature beats,etc.can all be diagnosed by ECG.When the distribution of training data and test data is inconsistent,the accuracy of the model will be affected.This phenomenon is called dataset shift.In the real-world heartbeat classification system,the heartbeat of the training set and test set often comes from patients of different ages and genders,so there are differences in the distribution of data sets.The main challenge in applying machine learning algorithms to clinical AI systems is dataset shift.Test-time adaptation(TTA)aims to adapt a pre-trained model from the source domain(SD)to the target domain(TD)without using any SD data or TD labels,thereby reducing model performance degradation due to domain differences.We propose a method based on multimodal image fusion and continual test-time adaptation(FCTA)for accurate and efficient heartbeat classification.First,the original ECG data is converted into a three-channel color image through a multimodal image fusion framework.The impact of class imbalance on network performance is overcome using a batch weight loss function,and then the pretrained source model is adapted to the TD using a continual test-time adaptation(CTA)method.Although our method is very simple,compared with other domain adaptation methods,it can significantly improve model performance on the test set and reduce the impact caused by the difference in domain distribution.
基金supported by the National Postdoctoral Program for Innovative Talents(Grant No.BX20230215)China Postdoctoral Science Foundation(Grant No.2023M732219)Shanghai Municipal Science and Technology Major Project(Grant No.2021SHZDZX0102)。
文摘Arrhythmias may lead to sudden cardiac death if not detected and treated in time.A supraventricular premature beat(SPB)and premature ventricular contraction(PVC)are important categories of arrhythmia disease.Recently,deep learning methods have been applied to the PVC/SPB heartbeats detection.However,most researchers have focused on time-domain information of the electrocardiogram and there has been a lack of exploration of the interpretability of the model.In this study,we design an interpretable and accurate PVC/SPB recognition algorithm,called the interpretable multilevel wavelet decomposition deep network(IMWDDN).Wavelet decomposition is introduced into the deep network and the squeeze and excitation(SE)-Residual block is designed for extracting time-domain and frequency-domain features.Additionally,inspired by the idea of residual learning,we construct a novel loss function for the constant updating of the multilevel wavelet decomposition parameters.Finally,the IMWDDN is evaluated on the Third China Physiological Signal Challenge Dataset and the MIT-BIH Arrhythmia database.The comparison results show IMWDDN has better detection performance with 98.51%accuracy and a 93.75%F1-macro on average,and its areas of concern are similar to those of an expert diagnosis to a certain extent.Generally,the IMWDDN has good application value in the clinical screening of PVC/SPB heartbeats.
