Feature extraction for murmur detection based on support vector regression of time-frequency representations

Jaramillo-Garzón, Jorge Alberto; Quiceno-Manrique, A. F.; Godino-Llorente, Juan Ignacio; Castellanos-Domínguez, G.

doi:10.1109/iembs.2008.4649484

Cited by 7 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since then, many articles have been published on the PCG segmentation techniques, features selection, and classification methods. The proposed classification algorithms include logistic regression [ 15 ], random forest [ 16 ], K-nearest neighbours (KNN) [ 15 , 17 , 18 ], regression tree [ 19 ], support vector machine (SVM) [ 15 , 20 , 21 ], hidden Markov model (HMM) [ 22 ], and ANN and its variants [ 23 , 24 , 25 ]. However, it was almost impossible to systematically and uniformly evaluate and compare the early research performance in this field, as they used different datasets with different classification tasks.…”

Section: Related Workmentioning

confidence: 99%

The Effect of Signal Duration on the Classification of Heart Sounds: A Deep Learning Approach

Bao

Kamavuako

2022

Sensors

View full text Add to dashboard Cite

Deep learning techniques are the future trend for designing heart sound classification methods, making conventional heart sound segmentation dispensable. However, despite using fixed signal duration for training, no study has assessed its effect on the final performance in detail. Therefore, this study aims at analysing the duration effect on the commonly used deep learning methods to provide insight for future studies in data processing, classifier, and feature selection. The results of this study revealed that (1) very short heart sound signal duration (1 s) weakens the performance of Recurrent Neural Networks (RNNs), whereas no apparent decrease in the tested Convolutional Neural Network (CNN) model was found. (2) RNN outperformed CNN using Mel-frequency cepstrum coefficients (MFCCs) as features. There was no difference between RNN models (LSTM, BiLSTM, GRU, or BiGRU). (3) Adding dynamic information (∆ and ∆²MFCCs) of the heart sound as a feature did not improve the RNNs’ performance, and the improvement on CNN was also minimal (≤2.5% in MAcc). The findings provided a theoretical basis for further heart sound classification using deep learning techniques when selecting the input length.

show abstract

Section: Related Workmentioning

confidence: 99%

The Effect of Signal Duration on the Classification of Heart Sounds: A Deep Learning Approach

Bao

Kamavuako

2022

Sensors

View full text Add to dashboard Cite

show abstract

“…The analysis pipeline of the vast majority of all the CAD systems that Denoising techniques (step 1) range from linear filters (such as a combination of lowpass and high-pass filters) [12][13][14][15], wavelets-based methods [15,16] and Kalman filters [17]; signal segmentation includes include ML-based methods [15], while feature extraction leverages empirical mode decomposion (EMD) [18] or information theory-based methods (such as entropy and Shannon-information) [19] and hidden Markov models [20]. Techniques employed for feature extraction (step 2) range from discrete cosine transforms (DCT) [21], discrete wavelet transforms (DWT) [22][23][24], short-time Fourier transforms (STFT) [21,25], mel-frequency cepstrum coefficients (MFCC) [26,27] and Choi-Williams distributions (CWD) [25]. Finally, the classification step ranges from relatively simple techniques like support vector machines (SVMs) [24,26,28], K-nearest neighbour (KNN) [25], artificial neural networks architectures [19,29,30] or a combination of such techniques, such as Adaptive Boosting in conjuction with Convolutional Neuronal Network (CNN) [31].…”

Section: (B) Automated Phonocardiogram Processing For Computer-aided Diagnosismentioning

confidence: 99%

“…Techniques employed for feature extraction (step 2) range from discrete cosine transforms (DCT) [21], discrete wavelet transforms (DWT) [22][23][24], short-time Fourier transforms (STFT) [21,25], mel-frequency cepstrum coefficients (MFCC) [26,27] and Choi-Williams distributions (CWD) [25]. Finally, the classification step ranges from relatively simple techniques like support vector machines (SVMs) [24,26,28], K-nearest neighbour (KNN) [25], artificial neural networks architectures [19,29,30] or a combination of such techniques, such as Adaptive Boosting in conjuction with Convolutional Neuronal Network (CNN) [31]. See [32] for a general overview.…”

Section: (B) Automated Phonocardiogram Processing For Computer-aided Diagnosismentioning

confidence: 99%

A novel multi-branch architecture for state of the art robust detection of pathological phonocardiograms

Duggento

Conti

Guerrisi

et al. 2021

Phil. Trans. R. Soc. A.

View full text Add to dashboard Cite

Heart auscultation is an inexpensive and fundamental technique to effectively diagnose cardiovascular disease. However, due to relatively high human error rates even when auscultation is performed by an experienced physician, and due to the not universal availability of qualified personnel, e.g. in developing countries, many efforts are made worldwide to propose computational tools for detecting abnormalities in heart sounds. The large heterogeneity of achievable data quality and devices, the variety of possible heart pathologies, and a generally poor signal-to-noise ratio make this problem very challenging. We present an accurate classification strategy for diagnosing heart sounds based on (1) automatic heart phase segmentation, (2) state-of-the art filters drawn from the field of speech synthesis (mel-frequency cepstral representation) and (3) an ad hoc multi-branch, multi-instance artificial neural network based on convolutional layers and fully connected neuronal ensembles which separately learns from each heart phase hence implicitly leveraging their different physiological significance. We demonstrate that it is possible to train our architecture to reach very high performances, e.g. an area under the curve of 0.87 or a sensitivity of 0.97. Our machine-learning-based tool could be employed for heartsound classification, especially as a screening tool in a variety of situations including telemedicine applications. This article is part of the theme issue ‘Advanced computation in cardiovascular physiology: new challenges and opportunities’.

show abstract

“…The first published study on automatic heart sound classification can be traced back to 1963 when research used threshold to identify rheumatic heart disease [5]. Afterwards, more Machine Learning (ML) techniques have been explored, such as logistic regression [6], regression tree [7], K-nearest neighbours (KNN) [6,8,9], random forest [10], support vector machine (SVM) [6,11,12], hidden Markov model (HMM) [13], etc. Besides traditional ML approaches, neural network (NN) and its variants have also been applied to heart sound classification [14][15][16].…”

Section: Introductionmentioning

confidence: 99%

Time-Frequency Distributions of Heart Sound Signals: A Comparative Study using Convolutional Neural Networks

Bao¹,

Yan²,

Lam³

et al. 2022

Preprint

View full text Add to dashboard Cite

Time-Frequency Distributions (TFDs) support the heart sound characterisation and classification in early cardiac screening. However, despite the frequent use of TFDs in signal analysis, no study comprehensively compared their performances on deep learning for automatic diagnosis. Furthermore, the combination of signal processing methods as inputs for Convolutional Neural Networks (CNNs) has been proved as a practical approach to increasing signal classification performance. Therefore, this study aimed to investigate the optimal use of TFD/ combined TFDs as input for CNNs. The presented results revealed that: 1) The transformation of the heart sound signal into the TF domain achieves higher classification performance than using of raw signals. Among the TFDs, the difference in the performance was slight for all the CNN models (within 1.3% in average accuracy). However, Continuous wavelet transform (CWT) and Chirplet transform (CT) outperformed the rest. 2) The appropriate increase of the CNN capacity and architecture optimisation can improve the performance, while the network architecture should not be overly complicated. Based on the ResNet or SEResNet family results, the increase in the number of parameters and the depth of the structure do not improve the performance apparently. 3) Combining TFDs as CNN inputs did not significantly improve the classification results. The findings of this study provided the knowledge for selecting TFDs as CNN input and designing CNN architecture for heart sound classification.

show abstract

Feature extraction for murmur detection based on support vector regression of time-frequency representations

Cited by 7 publications

References 12 publications

The Effect of Signal Duration on the Classification of Heart Sounds: A Deep Learning Approach

The Effect of Signal Duration on the Classification of Heart Sounds: A Deep Learning Approach

A novel multi-branch architecture for state of the art robust detection of pathological phonocardiograms

Time-Frequency Distributions of Heart Sound Signals: A Comparative Study using Convolutional Neural Networks

Contact Info

Product

Resources

About