ECG Classification Using Wavelet Packet Entropy and Random Forests

Abstract

Abstract:The electrocardiogram (ECG) is one of the most important techniques for heart disease diagnosis. Many traditional methodologies of feature extraction and classification have been widely applied to ECG analysis. However, the effectiveness and efficiency of such methodologies remain to be improved, and much existing research did not consider the separation of training and testing samples from the same set of patients (so called inter-patient scheme). To cope with these issues, in this paper, we propose a method to classify ECG signals using wavelet packet entropy (WPE) and random forests (RF) following the Association for the Advancement of Medical Instrumentation (AAMI) recommendations and the inter-patient scheme. Specifically, we firstly decompose the ECG signals by wavelet packet decomposition (WPD), and then calculate entropy from the decomposed coefficients as representative features, and finally use RF to build an ECG classification model. To the best of our knowledge, it is the first time that WPE and RF are used to classify ECG following the AAMI recommendations and the inter-patient scheme. Extensive experiments are conducted on the publicly available MIT-BIH Arrhythmia database and influence of mother wavelets and level of decomposition for WPD, type of entropy and the number of base learners in RF on the performance are also discussed. The experimental results are superior to those by several state-of-the-art competing methods, showing that WPE and RF is promising for ECG classification.

A CEEMDAN and XGBOOST-Based Approach to Forecast Crude Oil Prices

“…According to (21), as for the tree structure q, the loss function at the leaf node j can be changed as…”

“…And then the results of all subtasks are accumulated as the final result. Based on this idea, a "decomposition and ensemble" framework was proposed and widely applied to the analysis of time series, such as energy forecasting [16,17], fault diagnosis [18][19][20], and biosignal analysis [21][22][23]. This framework consists of three stages.…”

Re-examination of Bregman functions and new properties of their divergences

“…This inequality and Proposition 5.1(II) imply that b is strongly convex on X with µ as a strong convexity parameter. From (75) and the fact that A is continuous and hence bounded, it follows that b ′ is Lipschitz continuous (and hence uniformly continuous) on X. To see that b ′ is weak-to-weak * sequentially continuous, one simply observes that if (x i ) ∞ i=1 is a sequence in X which converges weakly to x ∈ X, then this assumption, the fact that A = A * and the fact that the inner product is symmetric show that for every w ∈ X,…”

Heartbeat classification in wearables using multi-layer perceptron and time-frequency joint distribution of ECG

“…A deep neural network based classifier is used as an alternative in [8]. Other alternatives include the random forest approach of [9] and the decision tree approach of [10]. All these approaches use ECG time samples.…”

“…Significant research have been conducted recently to autonomously classify ECG heartbeats (QRS peaks) [3][4][5][6][7][8][9][10][11][12] (Section 2 provides details of these approaches.). These studies have the following limitations that we address in this work.…”