Kurtosis-based noise estimation and multiscale energy to denoise ECG signal

Sharma, L. N.; Dandapat, Samarendra; Mahanta, Anil

doi:10.1007/s11760-011-0227-7

Cited by 16 publications

(11 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the filtering part, the artifacts such as base line wandering and drift are filtered out by a moving average filter [20]. To remove high frequency noise, relative energies of wavelet subbands and a noise variance based threshold are used [21]. The frame based processing of 12-lead ECG can capture the intrarhythm, inter-sample and inter-lead correlation information.…”

Section: Methodsmentioning

confidence: 99%

“…The diagnostic information of an ECG signal are distributed in different wavelet subbands based upon their bandwidth or frequency content. It has been reported that the lower frequency subbands contain most of the diagnostically significant information of the ECG signal [23], [21]. If all 12-standard ECG leads are decomposed with same mother wavelet and decomposition levels, it results in similar subbands with equal number of coefficients.…”

Section: A Proposed Multiscale Feature Analysismentioning

confidence: 99%

“…In this work, the multiscale energy features from cA 6,k , cD 6,k , cD 5,k and cD 4,k subbands are evaluated. It is seen that the 'PQRST' morphologies and its changes can be captured by multiscale wavelet energies [21], [23]. The within-class variations of normalized multiscale wavelet energy for all ECG leads are evaluated using 1074 myocardial infarction and 1074 healthy control multilead ECG frames.…”

Section: A Proposed Multiscale Feature Analysismentioning

confidence: 99%

“…The orthonormal eigenvectors represent the signals in the directions of maximum variances. Most of the energies are retained in wavelet subbands, cA m 6,k , cD m 6,k , cD m 5,k and cD m 4,k , where m is ECG lead number [21], [23]. From each of the multiscale matrices, A 6 , D 6 , D 5 and D 4 , six dominant eigenvalues are used as features.…”

Section: A Proposed Multiscale Feature Analysismentioning

confidence: 99%

See 3 more Smart Citations

Multiscale Energy and Eigenspace Approach to Detection and Localization of Myocardial Infarction

Sharma

Tripathy

Dandapat

2015

IEEE Trans. Biomed. Eng.

Self Cite

262

140

View full text Add to dashboard Cite

In this paper, a novel technique on a multiscale energy and eigenspace (MEES) approach is proposed for the detection and localization of myocardial infarction (MI) from multilead electrocardiogram (ECG). Wavelet decomposition of multilead ECG signals grossly segments the clinical components at different subbands. In MI, pathological characteristics such as hypercute T-wave, inversion of T-wave, changes in ST elevation, or pathological Q-wave are seen in ECG signals. This pathological information alters the covariance structures of multiscale multivariate matrices at different scales and the corresponding eigenvalues. The clinically relevant components can be captured by eigenvalues. In this study, multiscale wavelet energies and eigenvalues of multiscale covariance matrices are used as diagnostic features. Support vector machines (SVMs) with both linear and radial basis function (RBF) kernel and K-nearest neighbor are used as classifiers. Datasets, which include healthy control, and various types of MI, such as anterior, anteriolateral, anterioseptal, inferior, inferiolateral, and inferioposterio-lateral, from the PTB diagnostic ECG database are used for evaluation. The results show that the proposed technique can successfully detect the MI pathologies. The MEES approach also helps localize different types of MIs. For MI detection, the accuracy, the sensitivity, and the specificity values are 96%, 93%, and 99% respectively. The localization accuracy is 99.58%, using a multiclass SVM classifier with RBF kernel.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: A Proposed Multiscale Feature Analysismentioning

confidence: 99%

Section: A Proposed Multiscale Feature Analysismentioning

confidence: 99%

Section: A Proposed Multiscale Feature Analysismentioning

confidence: 99%

See 2 more Smart Citations

Multiscale Energy and Eigenspace Approach to Detection and Localization of Myocardial Infarction

Sharma

Tripathy

Dandapat

2015

IEEE Trans. Biomed. Eng.

Self Cite

262

140

View full text Add to dashboard Cite

show abstract

“…In general, wavelet based denoising can be accomplished by thresholding coefficients with an appropriate method. In wavelet subbands, using higher order statistics, noise can be effectively removed [15], [16]. For denoising of multilead ECG signals, multiscale principal component analysis (MSPCA) is successfully applied [17], [18], [19].…”

Section: Power Line Interferencementioning

confidence: 99%

Denoising pathological multilead electrocardiogram signals using multiscale singular value decomposition

Sharma

2014

2014 Twelfth International Conference on ICT and Knowledge Engineering

Self Cite

View full text Add to dashboard Cite

In this paper, denoising of multilead electrocardiograms (ECG) using multiscale singular value decomposition is proposed. If signal of each ECG leads are wavelet transformed with same mother wavelet and decomposition levels, it helps formation of multivariate multiscale matrices at wavelet scales. Singular value decomposition is applies in these scales. A new method to select singular values at these scales is proposed which is based on weighted ratio of matrix norms. This optimizes the approximate ranks for multiscale multivariate matrices to capture the diagnostic components present at different scales. Testing with records from PTB diagnostic ECG database for various pathological cases gives better SNR improvement retaining the pathological signatures. After adding white Gaussian noise at different SNR levels, quantitative analysis is carried out by evaluating error measures like percentage root mean square difference (PRD), root mean square error (NRMSE) and wavelet energy based diagnostic distortion measure (WEDD).

show abstract

Detection of myocardial infarction from vectorcardiogram using relevance vector machine

Tripathy

Dandapat

2017

SIViP

View full text Add to dashboard Cite

Kurtosis-based noise estimation and multiscale energy to denoise ECG signal

Cited by 16 publications

References 19 publications

Multiscale Energy and Eigenspace Approach to Detection and Localization of Myocardial Infarction

Multiscale Energy and Eigenspace Approach to Detection and Localization of Myocardial Infarction

Denoising pathological multilead electrocardiogram signals using multiscale singular value decomposition

Detection of myocardial infarction from vectorcardiogram using relevance vector machine

Contact Info

Product

Resources

About