Classification of Four Types of Common Murmurs using Wavelets and a Learning Vector Quantization Network

Rios-Gutiérrez, Fernando; Alba-Flores, Rocio; Ejaz, K.; Nordehn, Glenn; Andrisevic, N.; Burns, S.G.

doi:10.1109/ijcnn.2006.1716385

Cited by 4 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the power of murmurs spreads across various frequency regions (0-400Hz), in order to examine the dominance of spectral power at specific frequencies, spectral power is computed in four frequency bands: 0-100Hz; 100-200Hz, 200-300Hz, and 300-400Hz. Hence, 4 features (spectral power [1][2][3][4], corresponding to the power in each of 4 bands, are computed by summing over frequency.…”

Section: ) Time Domain Featuresmentioning

confidence: 99%

“…In the past few years many researchers have attempted to solve this problem, having achieved significant results. Namely, a feedforward neural network for murmur detection and two Learning Vector Quantization (LVQ) networks were employed for murmur classification in [3]. The murmur classification system is designed to categorize the detected murmur into 3 common types: Mitral Stenosis, Aortic insufficiency and Mitral insufficiency.…”

Section: Introductionmentioning

confidence: 99%

“…In time domain, 9 values of Shannon energy (Shannon energy[1][2][3][4][5][6][7][8][9] are calculated. In the time-frequency domain 11 features are extracted using the Shannon entropy of the 10th level wavelet detail coefficients (WT entropy 1-11), 9 energy measures are obtained from the high frequency components of the signal using 6th level details wavelet coefficients (WT detail[1][2][3][4][5][6][7][8][9], 16 features are obtained from the S-transform (ST map 1-16), 8 eigenvalues are calculated from singular value decomposition of the S-transform, 40 features are extracted from two left and two right eigenvectors and their corresponding histograms are also included(Eigenvalues 1-8, Eigenfrequency 1 1-10, Eigenfrequency 2 1-10, Eigentime1 1-10, Eigentime2, 1-10).In the category of nonlinear and chaos based or state space based features, the applied feature set includes: 16 discretized equally sized triangles from non-redundant region of the bispectrum (HOS 1-16), 40 features from Gaussian mixture model of the reconstructed space of the heart cycle (GMMx cycle 1-40), 40 from murmur (GMMx murmur 1-40), 8 features obtained using the variance fractal dimension (VFD 1-8) and 10 features from recurrence quantification analysis (RQA…”

mentioning

confidence: 99%

See 2 more Smart Citations

Heart murmur classification with feature selection

Kumar

Carvalho

Antunes

et al. 2010

2010 Annual International Conference of the IEEE Engineering in Medicine and Biology

View full text Add to dashboard Cite

Heart sounds entail crucial heart function information. In conditions of heart abnormalities, such as valve dysfunctions and rapid blood flow, additional sounds are heard in regular heart sounds, which can be employed in pathology diagnosis. These additional sounds, or so-called murmurs, show different characteristics with respect to cardiovascular heart diseases, namely heart valve disorders. In this paper, we present a method of heart murmur classification composed by three basic steps: feature extraction, feature selection, and classification using a nonlinear classifier. A new set of 17 features extracted in the time, frequency and in the state space domain is suggested. The features applied for murmur classification are selected using the floating sequential forward method (SFFS). Using this approach, the original set of 17 features is reduced to 10 features. The classification results achieved using the proposed method are compared on a common database with the classification results obtained using the feature sets proposed in two well-known state of the art methods for murmur classification. The achieved results suggest that the proposed method achieves slightly better results using a smaller feature set.

show abstract

Section: ) Time Domain Featuresmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Heart murmur classification with feature selection

Kumar

Carvalho

Antunes

et al. 2010

2010 Annual International Conference of the IEEE Engineering in Medicine and Biology

View full text Add to dashboard Cite

show abstract

“…Extracting the features of heart sound signals and performing quantitative analysis are helpful for the early screening of heart disease. Choi [10] decomposed heart sound signals to improve the accuracy of feature extraction efficiency of classification and recognition of heart sound, and Gutierrez et al [11] used discrete wavelet transform and short-time time-frequency transform to extract heart sounds' feature parameters; in addition, these researchers used a vectorization model to classify and recognize four kinds of common heart murmurs. Neural network pattern recognition and heart sound classification replicate doctors' auscultation and analysis mechanism.…”

Section: Introductionmentioning

confidence: 99%

Feature extraction and classification of heart sound using 1D convolutional neural networks

Liu

Zhao

et al. 2019

EURASIP J. Adv. Signal Process.

View full text Add to dashboard Cite

We proposed a one-dimensional convolutional neural network (CNN) model, which divides heart sound signals into normal and abnormal directly independent of ECG. The deep features of heart sounds were extracted by the denoising autoencoder (DAE) algorithm as the input feature of 1D CNN. The experimental results showed that the model using deep features has stronger anti-interference ability than using mel-frequency cepstral coefficients, and the proposed 1D CNN model has higher classification accuracy precision, higher F-score, and better classification ability than backpropagation neural network (BP) model. In addition, the improved 1D CNN has a classification accuracy rate of 99.01%.

show abstract