ECG Heartbeat Classification: A Deep Transferable Representation

Abstract

Electrocardiogram (ECG) can be reliably used as a measure to monitor the functionality of the cardiovascular system. Recently, there has been a great attention towards accurate categorization of heartbeats. While there are many commonalities between different ECG conditions, the focus of most studies has been classifying a set of conditions on a dataset annotated for that task rather than learning and employing a transferable knowledge between different tasks. In this paper, we propose a method based on deep convolutional neural networks for the classification of heartbeats which is able to accurately classify five different arrhythmias in accordance with the AAMI EC57 standard. Furthermore, we suggest a method for transferring the knowledge acquired on this task to the myocardial infarction (MI) classification task. We evaluated the proposed method on PhysionNet's MIT-BIH and PTB Diagnostics datasets. According to the results, the suggested method is able to make predictions with the average accuracies of 93.4% and 95.9% on arrhythmia classification and MI classification, respectively.

Inter- and Intra- Patient ECG Heartbeat Classification for Arrhythmia Detection: A Sequence to Sequence Deep Learning Approach

“…All network weights were updated by the RMSProp algorithm with mini batches of size 20 and a learning rate of α = 0.001. [17] 93…”

“…These shallow machine learning methods for ECG processing usually follow three main steps, including 1) signal pre-processing, which includes noise removal methods, heartbeat segmentation, etc; 2) feature extraction; and 3) learning/classification. Even though these methods with hand-engineered features and applying noise removal techniques have achieved acceptable performances, deep learning approaches (i.e., automated feature extractions) have shown impressive results in various domains ranging from computer vision and reinforcement learning to natural language processing [9], [10], [11], [12], [13], [14], [15], [16] as well as more applicable outcomes in biomedical signal processing [17], [18], [19], [20].…”

Interpretable and Steerable Sequence Learning via Prototypes

“…The result is summarized in Table 3. We can see that ProSeNet has comparable performance to LSTM, and even slightly better accuracy than Residual CNN [16]. Our model can present verifiable and understandable prototypes which are very useful in the healthcare domain.…”

“…Recently, variants of RNNs including Long-Short Term Memory networks (LSTMs) [21] have been proven to be very effective in modeling sequence data. They have been successfully applied to sentiment analysis [34], ECG signal classification [16], mortality and disease risk prediction using EHR data [8,10,13], and etc..…”

“…In the dataset, each signal consists of heartbeats annotated by at least two cardiologists. We downsample the ECG signals to 125Hz and split them into annotated heartbeats according to the protocol proposed by Kachuee et al [16]. The annotations are mapped into five groups as suggested by AAMI [2]: Normal (N), Supraventricular Ectopic Beat (SVEB), Ventricular Ectopic Beat (VEB), Fusion Beat (F) and Unknown Beat (Q).…”