1. Introduction
Cardiovascular diseases (CVDs) are the foremost cause of human death worldwide, which can lead to over 31% of deaths every year. With the progressive aging of populations worldwide, the number of patients with CVDs may continue to increase. It is estimated that the number of deaths due to CVDs will increase from 17 million in 2016 to 24 million in 2030 [
1]. Therefore, monitoring and preventing CVDs in advance has become one of the important tasks for many countries [
2].
Arrhythmia is a common CVDs, which refers to a series of rhythm and/or waveform irregular. As one of the most common arrhythmias, premature ventricular contraction (PVC) is caused by premature ectopic beats in the right or left ventricle [
3]. Frequent PVC and multisource PVC detection have important clinical significance [
4]. Clinicians generally detect PVC by observing rhythmic changes and subtle morphological changes from electrocardiogram (ECG) signal. However, this visual inspection may increase the manual interpretation work for physicians and lead to low efficiency for long-term PVC recognition. In order to reduce the workload of clinicians and improve PVC detection accuracy, researchers developed computer-aided systems for automagical diagnosis [
5].
Various automatic ECG heartbeat classification algorithms have been developed in recent decades, which can be summarized into two categories: expert system (ES)-based and deep learning (DL)-based methods. The ES-based methods classify heartbeats into different categories by judging multiple features with fixed thresholds. Most ES-based algorithms utilize rule-based features derived from rhythmic intervals (RR-interval, QT-interval, PR-interval, etc.) and morphological characteristics (P-wave, Q-wave, T-wave, etc.). Liu et al. [
6] presented a personalized ECG template construction method and detected PVC beats based on template matching, and the sensitivity (Se) on the MIT-BIH arrhythmia database (MIT-BIH-AR) (DS2) reached over 99%. Although this method has low computational complexity and can be applied for real-time conditions, the high performance is not tested on other databases especially on the dynamic noisy signals. Nahar et al. [
7] proposed an algorithm for PVC detection based on morphological transformation and cross-correlation technology, which used the morphological features to directly detect PVC. The potential of this proposed method was examined using 32 records from the MIT-BIH-AR database, reporting a specificity (Sp) of 96.67%, and a Se of 95.2%. Li et al. [
8] proposed a low-complexity data-adaptive approach for PVC recognition. They tested the method on INCART database and achieved a Se of 93.4%, an accuracy (ACC) of 94%, and a positive predictive value (P+) of 66.5%. These methods can be used for real-time applications without patient-specific consideration, as these methods have low computational complexity and good generalization capabilities. However, they need professional researchers to choose features and specific thresholds according to different tasks. Moreover, these detailed features are susceptible to noise interference, resulting in poor anti-noise ability of the algorithm.
With the development of machine learning, numerous DL-based methods have been developed, including auto-encoding (AE) [
9], convolutional neural network (CNN) [
10], block-based neural network (BBNN) [
11], long-short term memory (LSTM) [
12], support vector machine (SVM) [
13], decision tree [
14], cascade forward neural network (CFNN) [
15], and random forest [
16], etc. The DL-based method omits the handcrafted features extraction process, as the DL network can automatically extract the high-dimensional features. Therefore, DL-based methods can be applied in situations with big data processing capabilities, such as cloud computing platforms [
17]. Yildirim et al. [
1] presented a new 1D-convolutional neural network model for cardiac arrhythmia detection based on long-duration ECG signal analysis, which achieved an ACC of 91.33% for 17 cardiac arrhythmia classes classification in the MIT-BIH-AR database. Similarly, Pławiak et al. [
18] proposed genetic ensembles of SVM-based classifiers for the same classification task and achieved a Se of 91.40% and an ACC of 98.99%. These two methods can be used for real-time signal processing and cloud computing on mobile devices, as they eliminate the need for detection and segmentation of QRS complexes. However, neither of these two methods can classify ECG segments that contain multiple ECG abnormalities. Shadmand et al. [
11] employed the particle swarm optimization algorithm to optimize the structure and weights of BBNN and obtained an accuracy of 97.00% for five classes of ECG classification on the MIT-BIH-AR database. This method highly relied on large volumes of labeled data and computing resources to obtain its satisfactory performance on different databases.
Although the reported ES- and DL-based automatic heartbeat classification algorithms can achieve high performances on different databases, the extracted features of ES-based method require professional knowledge and are susceptible to noise; while the DL-based method is unexplainable and is easy to overfit on a small amount of labeled data. Therefore, in order to ensure the accuracy of ES-based and DL-based algorithms while considering the disadvantages of these two methods, a robust PVC identification algorithm based on a novel expert system and deep learning combination strategy was proposed in this paper. To evaluate its performance and generalization capacity, the method was tested on three different databases: the MIT-BIH-AR database, the St. Petersburg Institute of Cardiological Technics (INCART) database and the China Physiological Signal Challenge 2020 (CPSC2020) database. There are three major contributions of the proposed work. (1) This article proposed a novel expert system and deep learning combination strategy for PVC recognition in single-lead ECG. (2) The developed PVC detection algorithm is unsupervised, since the employed LSTM-AE network is used as the feature extraction process for heartbeat clustering. (3) The designed method is less complex and lightweight compared to most of the proposed automatic PVC detection methods.
5. Discussion
A PVC recognition algorithm based on integrating deep learning and rules was proposed in this study. Many ES-based or DL-based automatic ECG heartbeat classification algorithms have achieved high recognition results. However, they are complementary in terms of robustness and generalization.
The contribution of this paper is the combination of the DL-assisted template construction and ES-based heartbeat classification, which not only guarantees the accuracy but also improves the interpretability, robustness and generalization ability of the algorithm. A wavelet-based statistical process control (SPC) method was proposed for PVC recognition on MIT-BIH-AR database [
28], the overall ACC was 97.90%, and the Se and P+ for PVC were 87.20% and 84.60%, respectively. This method could improve PVC sensitivity by manually adjusting parameter thresholds according to different situations, while our method could achieve high PVC sensitivity without any manual process. A real-time premature beat (PB) detection method for single-lead ECG was proposed based on several simple rules [
26], which was reported to have low computational complexity and could be used for real-time PB detection for portable ambulatory ECG monitoring. However, their accuracy on the total data (85.56%) was still non-neglected for accurate clinical diagnosis. Malek et al. [
29] developed an improved template matching technique for identifying normal and PVC beats in ECG signals, which was evaluated on the INCART, QT, MIT-BIH Supraventricular Arrhythmia, and Fantasia databases, and the accuracy was 97.91%, 99.34%, 99.89%, and 98.44%, respectively. One of the strengths of this method was the application of an adaptable threshold without the need for expert intervention, however, the features they adopted were more complex than ours. Talbi et al. [
30] studied the effectiveness of the fractional linear prediction (FLP) technique on the ECG signal modeling, and developed a PVC recognition method based on the three coefficients of FLP and KNN, and the best accuracy of 96% was achieved on MIT-BIH-AR database. Most of the existing ES-based methods are efficient and requires less expert intervention, but the robustness still needs to be improved for daily life application.
From
Table 4, we compared the PVC recognition between the proposed method with existing methods on MIT-BIH-AR database and INCART database. The satisfactory performance of the proposed method on these two clinical databases demonstrated that our method not only guarantees the accuracy and robustness advantages of DL-based method, but also improved the generalization capacity and interpretability advantages of ES-based methods.
With the popularity of machine learning, many researchers have implemented machine learning algorithms in arrhythmia recognition and achieved high performance. Mazidi et al. [
32] designed a linear kernel-based SVM classifier with morphology, time domain, time-frequency domain and nonlinear features for PVC recognition, the method achieved a higher overall ACC and Se (99.78% and 99.91%, respectively) than our method. Wang et al. [
34] proposed a PVC detection scheme based on image processing and CNN for scanned clinical ECG reports, and their Se and ACC could reach 95.47% and 98.25%, respectively. However, our method was unsupervised while the training set used in their method was overlapped in their test set. Oh et al. [
12] proposed an automated system using a combination of CNN and LSTM for variable-length ECG classification (five class), they obtained the high classification accuracy of 98.10% without noise elimination on the MIT-BIH-AR database. The system could analyze ECG signals of different lengths with only a single type of arrhythmia, but it was computationally intensive. Yang et al. [
27] applied stacked sparse autoencoders (SSAEs) and a Softmax regression (SF) for six types of ECG classification and achieved average 99.22% Se and 99.37% P+ on MIT-BIH-AR database. The features extracted by SSAE had no individual independent differences in feature selection and extraction accuracy, and almost no useful heartbeat information was lost. However, the method was semisupervised and required trained cardiologists to first classify each beat cluster into normal or ventricular. Therefore, it was inappropriate for analyzing long-term signals.
Although we did not participate in CPSC2020 as we were affiliated with the organizer of the challenge, the performance of the proposed method on long-term wearable ECG database (CPSC2020) was also compared with the published top five teams for PVC recognition in CPSC2020 (
Table 3). The method proposed by the published champion team employed DenseNet model to classify the heartbeats into three categories (normal, premature ventricular contraction and supraventricular premature beat) and refined the results by a postprocessing procedure with several clinical rules. The algorithms of other teams were almost all DL-based methods, and they could achieve excellent performance on the training set, but they could not maintain such good results on the test set. The reason might be that these teams overoptimized the accuracy of their algorithm on the training set, leading to overfitting, which affected the algorithm results on hidden test set. Both our method and the published champion team’s results outperformed DL-based methods, indicating that the fusion of these two (ES-based and DL-based) methods had the potential to reform the existing methods based only on ES or DL.
To evaluate the computational complexity of our method, we computed and compared the operating time of our method and the CPSC2020 top five teams on the hidden test set. In addition, we also compared the running time with some published works in parallel. Three morphological features and seven statistical features were directly extracted, normalized and fed into CFNN classifier for PVC recognition, which could process 20-s segment within 2.1 s on a Samsung Galaxy J1 motherboard (a quad-core Cortex-A7 CPU clocked at up to 1.2 GHz with 1 GB RAM, OS Android 6.0) [
15]. Khalaf et al. [
37] proposed an SVM-based method on MATLAB R2010a on Intel
® Core™ i5 3.2 GHz processor and 8 GB RAM, and it consumed 54.8 ms for each beat classification. Arrais Junior et al. [
38] reported an adaptive threshold and redundant discrete wavelet transform fusion method, which can process 30 min signals using only 61.2 s on the Matlab 2014a platform. These results showed that (1) the superposition of deep learning and time-frequency conversion processes will increase the complexity of the algorithm; (2) complex deep learning frameworks are indeed more time-consuming than simple CNN; (3) the DL-based feature extraction + ES-based postprocessing analysis generally take less time. The comparison results further verified the advantage of the fusion of these two (ES-based and DL-based) methods.
The employed DL-based method (LSTM-AE module) was used to extract features from ECG heartbeats for K-means clustering, and the PVC identification was based on a combination of multiple rules, including template matching and rhythm characteristics. The features used for classification are extracted according to the R-peak-relevant clinical experience: the Covr, ArDiff and EnDiff are used to map the morphological and frequency domain difference between PVC and Non_PVC, and the rhythmic rules are used to map the variation of RR intervals between PVC and Non_PVC. All these features are extracted only based on R peaks instead of those complex features detected from precise fiducial points (Q wave, S wave, etc.) and professional knowledge, which can not only retain the interpretability of the proposed algorithm, but also improve the antinoise ability of the algorithm.
Although the proposed method is an important contribution to unsupervised PVC identification, there are three main limitations. (1) The performance is affected by the misalignment of QRS complex, more accurate QRS detection algorithm should be designed to detect the peak of each QRS complex for precise ECG classification. (2) This method is trained and tested only on the Windows platform, so further work is needed to embed the algorithm to the mobile terminal for daily life monitoring application. (3) Only one-channel information is considered in this paper, multichannel information should be considered from multilead ECG monitoring systems for accuracy improvement of PVC recognition, or even other kinds of heartbeat classification.