Limits...
A framework for automatic heart sound analysis without segmentation.

Yuenyong S, Nishihara A, Kongprawechnon W, Tungpimolrut K - Biomed Eng Online (2011)

Bottom Line: Geometric mean was used as performance index.Average classification performance using ten-fold cross-validation was 0.92 for noise free case, 0.90 under white noise with 10 dB signal-to-noise ratio (SNR), and 0.90 under impulse noise up to 0.3 s duration.Further work include building a new training set recorded from actual patients, then further evaluate the method based on this new training set.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Communication and Integrated Systems, Tokyo Institute of Technology, Japan 2-12-1-W9-108 Ookayama, Meguro-ku, Tokyo, Japan. toey123@gmail.com

ABSTRACT

Background: A new framework for heart sound analysis is proposed. One of the most difficult processes in heart sound analysis is segmentation, due to interference form murmurs.

Method: Equal number of cardiac cycles were extracted from heart sounds with different heart rates using information from envelopes of autocorrelation functions without the need to label individual fundamental heart sounds (FHS). The complete method consists of envelope detection, calculation of cardiac cycle lengths using auto-correlation of envelope signals, features extraction using discrete wavelet transform, principal component analysis, and classification using neural network bagging predictors.

Result: The proposed method was tested on a set of heart sounds obtained from several on-line databases and recorded with an electronic stethoscope. Geometric mean was used as performance index. Average classification performance using ten-fold cross-validation was 0.92 for noise free case, 0.90 under white noise with 10 dB signal-to-noise ratio (SNR), and 0.90 under impulse noise up to 0.3 s duration.

Conclusion: The proposed method showed promising results and high noise robustness to a wide range of heart sounds. However, more tests are needed to address any bias that may have been introduced by different sources of heart sounds in the current training set, and to concretely validate the method. Further work include building a new training set recorded from actual patients, then further evaluate the method based on this new training set.

Show MeSH
Cross-validation procedure where k is the number of repetitions that each NN is trained in each fold. SAE is the sum of absolute error defined in Equation 10. The process inside each box is the training for each network.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3045988&req=5

Figure 13: Cross-validation procedure where k is the number of repetitions that each NN is trained in each fold. SAE is the sum of absolute error defined in Equation 10. The process inside each box is the training for each network.

Mentions: Ten-fold cross validation that incorporates bagging classifiers was used as performance evaluation; the flowchart of which is shown in Figure 13. It consists of dividing the data into ten portions. Then, in each fold, one portion was reserved for testing while the remaining nine became the training set, which was sampled according to the bootstrapping procedure discussed in section 2.5. This generated N different training sets for each of the N neural networks. The networks were trained and their performance was evaluated on the test portion, which is the same for all networks since the test portion was not sampled. Each network was trained 30 times and the instances that yielded the best performance were kept. This procedure is illustrated in Figure 13 inside the bounding boxes. Each box represents the training process of a single network. A network's performance was evaluated by the sum of absolute error (SAE), defined in Equation 10:


A framework for automatic heart sound analysis without segmentation.

Yuenyong S, Nishihara A, Kongprawechnon W, Tungpimolrut K - Biomed Eng Online (2011)

Cross-validation procedure where k is the number of repetitions that each NN is trained in each fold. SAE is the sum of absolute error defined in Equation 10. The process inside each box is the training for each network.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3045988&req=5

Figure 13: Cross-validation procedure where k is the number of repetitions that each NN is trained in each fold. SAE is the sum of absolute error defined in Equation 10. The process inside each box is the training for each network.
Mentions: Ten-fold cross validation that incorporates bagging classifiers was used as performance evaluation; the flowchart of which is shown in Figure 13. It consists of dividing the data into ten portions. Then, in each fold, one portion was reserved for testing while the remaining nine became the training set, which was sampled according to the bootstrapping procedure discussed in section 2.5. This generated N different training sets for each of the N neural networks. The networks were trained and their performance was evaluated on the test portion, which is the same for all networks since the test portion was not sampled. Each network was trained 30 times and the instances that yielded the best performance were kept. This procedure is illustrated in Figure 13 inside the bounding boxes. Each box represents the training process of a single network. A network's performance was evaluated by the sum of absolute error (SAE), defined in Equation 10:

Bottom Line: Geometric mean was used as performance index.Average classification performance using ten-fold cross-validation was 0.92 for noise free case, 0.90 under white noise with 10 dB signal-to-noise ratio (SNR), and 0.90 under impulse noise up to 0.3 s duration.Further work include building a new training set recorded from actual patients, then further evaluate the method based on this new training set.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Communication and Integrated Systems, Tokyo Institute of Technology, Japan 2-12-1-W9-108 Ookayama, Meguro-ku, Tokyo, Japan. toey123@gmail.com

ABSTRACT

Background: A new framework for heart sound analysis is proposed. One of the most difficult processes in heart sound analysis is segmentation, due to interference form murmurs.

Method: Equal number of cardiac cycles were extracted from heart sounds with different heart rates using information from envelopes of autocorrelation functions without the need to label individual fundamental heart sounds (FHS). The complete method consists of envelope detection, calculation of cardiac cycle lengths using auto-correlation of envelope signals, features extraction using discrete wavelet transform, principal component analysis, and classification using neural network bagging predictors.

Result: The proposed method was tested on a set of heart sounds obtained from several on-line databases and recorded with an electronic stethoscope. Geometric mean was used as performance index. Average classification performance using ten-fold cross-validation was 0.92 for noise free case, 0.90 under white noise with 10 dB signal-to-noise ratio (SNR), and 0.90 under impulse noise up to 0.3 s duration.

Conclusion: The proposed method showed promising results and high noise robustness to a wide range of heart sounds. However, more tests are needed to address any bias that may have been introduced by different sources of heart sounds in the current training set, and to concretely validate the method. Further work include building a new training set recorded from actual patients, then further evaluate the method based on this new training set.

Show MeSH