Journal Article•
PCA and ICA applied to noise reduction in multi-lead ECG
IMEC1
TL;DR: The performance of PCA and ICA in the context of cleaning noisy ECGs in ambulatory conditions was investigated in this article, where the output of a beat detection algorithm was applied to both the output signal after PCA/ICA filtering and compared to the detections in the signal before filtering.
Abstract: The performance of PCA and ICA in the context of cleaning noisy ECGs in ambulatory conditions was investigated. With this aim, ECGs with artificial motion artifacts were generated by combining clean 8-channel ECGs with 8-channel noise signals at SNR values ranging from 10 down to −10 dB. For each SNR, 600 different simulated ECGs of 10-second length were selected. 8-channel PCA and ICA were applied and then inverted after selecting a subset of components. In order to evaluate the performance of PCA and ICA algorithms, the output of a beat detection algorithm was applied to both the output signal after PCA/ICA filtering and compared to the detections in the signal before filtering. Applying both PCA and ICA and retaining the optimal component subset, yielded sensitivity (Se) of 100% for all SNR values studied. In terms of Positive predictivity (+P), applying PCA, yielded to an improvement for all SNR values as compared to no cleaning (+P=95.45% vs. 83.09% for SNR=0dB; +P=56.87% vs. 48.81% for SNR=−10dB). However, ICA filtering gave a higher improvement in +P for all SNR values (+P=100.00% for SNR=0dB; +P=61.38% for SNR=−10dB). An automatic method for selecting the components was proposed. By using this method, both PCA and ICA gave an improvement as compared to no filtering over all SNR values. ICA had a better performance (SNR=−5dB, improvement in +P of 8.33% for PCA and 22.92% for ICA).
Citations
More filters
TL;DR: The results show that the electrode-tissue impedance can correlate with the motion artifacts for local disturbance of the electrodes and that the impedance signals can be used in motion artifact removal techniques such as adaptive filtering.
Abstract: Ambulatory monitoring of the electrocardiogram (ECG) is a highly relevant topic in personal healthcare. A key technical challenge is overcoming artifacts from motion in order to produce ECG signals capable of being used in clinical diagnosis by a cardiologist. An electrode-tissue impedance is a signal of significant interest in reducing the motion artifact in ECG recordings on the go. A wireless system containing an ultralow-power analog front-end ECG signal acquisition, as well as the electrode-tissue impedance, is used in a validation study on multiple subjects. The goal of this paper is to study the correlation between motion artifacts and skin electrode impedance for a variety of motion types and electrodes. We have found that the correlation of the electrode-tissue impedance with the motion artifact is highly dependent on the electrode design the impedance signal (real, imaginary, absolute impedance), and artifact types (e.g., push or pull electrodes). With the chosen electrodes, we found that the highest correlation was obtained for local electrode artifacts (push, pull, electrode) followed by local skin (stretch, twist, skin) and global artifacts (walk, jog, jump). The results show that the electrode-tissue impedance can correlate with the motion artifacts for local disturbance of the electrodes and that the impedance signals can be used in motion artifact removal techniques such as adaptive filtering.
60 citations
Cites methods from "PCA and ICA applied to noise reduct..."
...have been used for separating MA and ECG [15], assuming that the sources of ECG and MA are uncorrelated....
[...]
Journal Article•
IMEC1
TL;DR: The performance of several adaptive filter (AdF) algorithm implementations was investigated in the context of cleaning noisy ambulatory ECGs and using AdF algorithm improved the performance of a beat detection (BD) algorithm as compared to non-filtering.
Abstract: The performance of several adaptive filter (AdF) algorithm implementations was investigated in the context of cleaning noisy ambulatory ECGs. Together with a noisy ECG signal, both body movement measured with accelerometers and skin-electrode impedance (SEI) were considered as reference signals to the AdF. ECG with artificial motion artifacts were generated by combining clean ECGs with noise signals. Several implementations and combinations of AdFs, and two reference signals (accelerometers and SEI) were investigated. Performance was measured by evaluating the output (sensitivity (Se) and positive predictivity (+P)) of a beat detection (BD) algorithm. Using AdF algorithm improved the performance of a BD algorithm as compared to non-filtering. SEI used as reference signal outperformed accelerometers. A variant of LMS, LMS sign-error, gave the best performance from all implementations considered. However, distortion observed in the filtered signal is high and therefore, these results cannot be extended to other features within the ECG.
33 citations
Cites background from "PCA and ICA applied to noise reduct..."
...However, distortion observed in the filtered signal is high and therefore, these results cannot be extended to other features within the ECG....
[...]
TL;DR: This approach combines JADE source separation and binary decision tree for identification and subsequent ECG noise removal and noticeably significantly better results were achieved when uncommon noise (electrode cable movement artefact) were compared.
Abstract: We have developed a method focusing on ECG signal de-noising using Independent component analysis (ICA). This approach combines JADE source separation and binary decision tree for identification and subsequent ECG noise removal. In order to to test the efficiency of this method comparison to standard filtering a wavelet- based de-noising method was used. Freely data available at Physionet medical data storage were evaluated. Evaluation criteria was root mean square error (RMSE) between original ECG and filtered data contaminated with artificial noise. Proposed algorithm achieved comparable result in terms of standard noises (power line interference, base line wander, EMG), but noticeably significantly better results were achieved when uncommon noise (electrode cable movement artefact) were compared.
26 citations
TL;DR: RLS method is much more effective and powerful than other methods in ECG noise cancellation, and even better than UNANR, and the introduced optimized method with adaptive threshold value would have great potential in biomedical application of signal processing and other fields.
Abstract: The electrocardiogram (ECG) is generally used for the diagnosis of cardiovascular diseases. In many of the biomedical applications, it is necessary to remove the noise from ECG recordings. Several adaptive filter structures have been proposed for noise cancellation. Compared to the least mean square (LMS) method, the unbiased and normalized adaptive noise reduction (UNANR) algorithm has better performance, as mentioned in previous investigations. In this paper, we review various kinds of ECG noise reduction algorithms. To provide a detailed and fair comparison, all normalized LMS (NLMS), Block LMS (BLMS), recursive least squares (RLS) and UNANR algorithms are implemented and their performance have been assessed using the same dataset and compared to different state-of-the-art approaches. Then, the performance analysis of all five algorithms is presented and compared in term of mean squared error (MSE), computational complexity and stability. The obtained results revealed that RLS method is much more effective and powerful than other methods in ECG noise cancellation, and even better than UNANR. Then, in order to reach the best performance of the mentioned filter and also, to minimize the output signal error, the optimized parameters of the algorithm were defined and results were investigated. The obtained outcomes show that the best Lambda (λ) occurs between 0.05 and 0.9, so that the convergence rate of the optimized RLS filter is faster than others. It not only decreases the noise, but also the ECG waveform is better conserved. Furthermore, the introduced optimized method with adaptive threshold value would have great potential in biomedical application of signal processing and other fields.
26 citations
TL;DR: The most commonly used machine learning and feature mining tools and several new trends and tendencies such as deep learning and biological networks for computational biomedicine are analyzed.
Abstract: This survey paper attempts to cover a broad range of topics related to computational biomedicine. The field has been attracting great attention due to a number of benefits it can provide the society with. New technological and theoretical advances have made it possible to progress considerably. Traditionally, problems emerging in this field are challenging from many perspectives. In this paper, we considered the influence of big data on the field, problems associated with massive datasets in biomedicine and ways to address these problems. We analyzed the most commonly used machine learning and feature mining tools and several new trends and tendencies such as deep learning and biological networks for computational biomedicine.
25 citations
References
More filters
Book•
01 May 1986TL;DR: In this article, the authors present a graphical representation of data using Principal Component Analysis (PCA) for time series and other non-independent data, as well as a generalization and adaptation of principal component analysis.
Abstract: Introduction * Properties of Population Principal Components * Properties of Sample Principal Components * Interpreting Principal Components: Examples * Graphical Representation of Data Using Principal Components * Choosing a Subset of Principal Components or Variables * Principal Component Analysis and Factor Analysis * Principal Components in Regression Analysis * Principal Components Used with Other Multivariate Techniques * Outlier Detection, Influential Observations and Robust Estimation * Rotation and Interpretation of Principal Components * Principal Component Analysis for Time Series and Other Non-Independent Data * Principal Component Analysis for Special Types of Data * Generalizations and Adaptations of Principal Component Analysis
17,446 citations
15 Oct 2005
TL;DR: Principal component analysis (PCA) as discussed by the authors replaces the p original variables by a smaller number, q, of derived variables, the principal components, which are linear combinations of the original variables.
Abstract: When large multivariate datasets are analyzed, it is often desirable to reduce their dimensionality. Principal component analysis is one technique for doing this. It replaces the p original variables by a smaller number, q, of derived variables, the principal components, which are linear combinations of the original variables. Often, it is possible to retain most of the variability in the original variables with q very much smaller than p. Despite its apparent simplicity, principal component analysis has a number of subtleties, and it has many uses and extensions. A number of choices associated with the technique are briefly discussed, namely, covariance or correlation, how many components, and different normalization constraints, as well as confusion with factor analysis. Various uses and extensions are outlined.
Keywords:
dimension reduction;
factor analysis;
multivariate analysis;
variance maximization
14,773 citations
TL;DR: The basic theory and applications of ICA are presented, and the goal is to find a linear representation of non-Gaussian data so that the components are statistically independent, or as independent as possible.
8,231 citations
"PCA and ICA applied to noise reduct..." refers background in this paper
...ICA had a better performance (SNR=-5dB, improvement in +P of 8.33% for PCA and 22.92% for ICA)....
[...]
TL;DR: An R-wave detector is developed and tested using patient signals recorded in the Coronary Care Unit of the Royal Infirmary of Edinburgh and with the MIT/BIH database, achieving a sensitivity and positive predictive value of 99.73% and 99.68%, respectively.
Abstract: The problem of automatic beat recognition in the ECG is tackled using continuous wavelet transform modulus maxima (CWTMM). Features within a variety of ECG signals can be shown to correspond to various morphologies in the CWTMM domain. This domain has an easy interpretation and offers a useful tool for the automatic characterization of the different components observed in the ECG in health and disease. As an application of this enhanced time-frequency analysis technique for ECG signals, an R-wave detector is developed and tested using patient signals recorded in the Coronary Care Unit of the Royal Infirmary of Edinburgh (attaining a sensitivity of 99.53% and a positive predictive value of 99.73%) and with the MIT/BIH database (attaining a sensitivity of 99.70% and a positive predictive value of 99.68%).
125 citations
"PCA and ICA applied to noise reduct..." refers methods in this paper
...With this dimensional reduction, these techniques look for simplifying a statistical problem with the minimal loss of information....
[...]
23 Oct 2002
TL;DR: This paper presents initial results of a novel approach to reducing ECG motion artifact using electrode motion as the reference signal to an adaptive filter and the motion signal and shows that the induced motion artifact was reduced in all data sets.
Abstract: The electrocardiogram (ECG) is the body-surface manifestation of the electrical potentials produced by the heart. The ECG is acquired by placing electrodes on the patient's skin. Motion artifact is the noise that results from motion of the electrode in relation to the patient's skin. Motion artifact can produce large amplitude signals in the ECG that may be misinterpreted by clinicians and automated systems resulting in misdiagnosis, prolonged procedure duration, and delayed or inappropriate treatment decisions. Motion artifact reduction is an unsolved problem because its frequency spectrum overlaps that of the ECG. This paper presents initial results of a novel approach to reducing ECG motion artifact. The hypothesis is that motion artifact can be reduced using electrode motion as the reference signal to an adaptive filter. Electrode motion was measured with two custom-developed sensors that utilized anisotropic magnetoresistive sensors and accelerometers. Motion artifact was induced by manually pushing on the electrode, pushing on the skin around the electrode, and pulling on the lead wire. Using an adaptive filter and the motion signal, the induced motion artifact was reduced in all data sets.
121 citations
"PCA and ICA applied to noise reduct..." refers methods in this paper
...By using this method, both PCA and ICA gave an improvement as compared to no filtering over all SNR values....
[...]