In this paper, we developed and evaluated a robust single-lead electrocardiogram (ECG) delineation system based on the wavelet transform (WT). In a first step, QRS complexes are detected. Then, each QRS is delineated by detecting and identifying the peaks of the individual waves, as well as the complex onset and end. Finally, the determination of P and T wave peaks, onsets and ends is performed. We evaluated the algorithm on several manually annotated databases, such as MIT-BIH Arrhythmia, QT, European ST-T and CSE databases, developed for validation purposes. The QRS detector obtained a sensitivity of Se = 99.66% and a positive predictivity of P+ = 99.56% over the first lead of the validation databases (more than 980,000 beats), while for the well-known MIT-BIH Arrhythmia Database, Se and P+ over 99.8% were attained. As for the delineation of the ECG waves, the mean and standard deviation of the differences between the automatic and manual annotations were computed. The mean error obtained with the WT approach was found not to exceed one sampling interval, while the standard deviations were around the accepted tolerances between expert physicians, outperforming the results of other well known algorithms, especially in determining the end of T wave.
In this paper, a dynamic linear approach was used over QT and RR series measured by an automatic delineator, to explore the interactions between QT interval variability (QTV) and heart rate variability (HRV). A low-order linear autoregressive model allowed to separate and quantify the QTV fractions correlated and not correlated with HRV, estimating their power spectral density measures. Simulated series and artificial ECG signals were used to assess the performance of the methods, considering a respiratory-like electrical axis rotation effect and noise contamination with a signal-to-noise ratio (SNR) from 30 to 10 dB. The errors found in the estimation of the QTV fraction related to HRV showed a nonrelevant performance decrease from automatic delineation. The joint performance of delineation plus variability analysis achieved less than 20% error in over 75% of cases for records presenting SNRs higher than 15 dB and QT standard deviation higher than 10 ms. The methods were also applied to real ECG records from healthy subjects where it was found a relevant QTV fraction not correlated with HRV (over 40% in 19 out of 23 segments analyzed), indicating that an important part of QTV is not linearly driven by HRV and may contain complementary information.
In IntroductionMyocardial ischemia is reflected in the ECG by amplitude changes in the ST segment and the T wave. The conventional ST level measurement (typically obtained at J+60 ms) represents a local measurement which unfortunately is vulnerable to various noise sources such as baseline wander and muscle; in addition, the measurement is rendered even more difficult due to heart rate related repolarization changes. In order to obtain more robust measurements, additional information can be introduced by making use of information from previous beats. For example, signal averaging [1] and its many variants relies on the observation that ischemia-induced beat-to-beat changes in ECG morphology are relatively slow.Another approach is to analyze the complete repolarization waveform (STT complex) by means of linear expansions to get a global characterization of the repolarization waveform in each lead [2]. A desired property of the basis functions is that they should characterize the relevant features in a small subspace. Then, a few subset of expansion coefficients will characterize the dominant signal waveform. The trends defined by the subset of expansion coefficients reflect the main beat-to-beat evolution of repolarization changes, in the same way as the ST trends.In this study we explore the spatial information available in multichannel ECG recordings for use in repolarization analysis. The repolarization waveforms from all leads are jointly analyzed by a truncated linear expansion model where the basis functions are matrices. Our hypothesis is that multichannel expansions may achieve a better packing of the signal energy than single-channel (temporal) expansions, taking into account the joint spatio-temporal information of the repolarization process.Linear expansions is a well-known technique for signal analysis and modelling. It is based on the decomposition of the signal as a linear combination of simple and elementary basis functions which define a new signal representation domain [3]. The selection of the domain is a key factor and should be done according to the properties of the analyzed signal and the application. The optimal (in the mean squared error sense) linear and unitary transform for signal coding is the Karhunen-Loève transform (KLT) [4]. This transformation is data-dependent and it is estimated from a data training set. Two different kinds of training sets can be used for ECG analysis: a unique training set formed by a large number of signals containing a wide range of waveforms, or smaller patient-specific training sets. The latter are often much more homogeneous, providing a better energy packing performance. However, if the training signals are heavily contaminated by noise and only a small number of occurrences are available, the KL basis functions may be greatly affected by noise. This may be the case of exercise test recordings.The aim of this work is twofold. Firstly, to introduce a multichannel signal model based on linear expansions which gives a joint spatio-temporal descript...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.