Article
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

This paper presents a multi-centroid diastolic duration model for the hidden semi-Markov model (HSMM) based heart sound segmentation. The centroids are calculated by hierarchical agglomerative clustering of the neighboring diastolic duration values using Ward’s method until center of clusters are found at least a systolic duration apart. The multiple peak distribution yields a sharper gradient of likelihood around the expected centroids and improves the discriminability of similar observations. The peak density at each centroid acts as a reference point for the HSMM to determine the origin of the hidden-state and adjust the corresponding state duration based on the maximum likelihood criterion. This model overcomes the limitation of the single peak mean value model that may overfit the duration distribution when the heart rate variation is relatively large. An extended logistic regression-HSMM algorithm using the proposed duration model is presented for the heart sound segmentation. In addition, the total variation filter is used to attenuate the effect of noises and emphasize the fundamental heart sounds, S1 and S2. The proposed method is evaluated on the training-set-a of 2016 Physionet/Computing in Cardiology Challenge and yields an average F1 score of 98.36 ± 0.43.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Nevertheless, the authors of reference [110] and [112] that used KNCN classifier achieved higher normal and abnormal PCG signal classification accuracy than the classification accuracy of reference [111], which only used KNN classification algorithm in PCG signal classification. A multi centroid diastolic duration model which is based on the hidden semi -Markov model (HSMM) was proposed in reference [113] to classify normal and abnormal PCG signals. The distance from the query points to the calculated centroids of the clusters are computed to determine their corresponding duration that can verify normal and abnormal states. ...
... As of the past research studies of [110], [111], [112], and [113], the heart sound signal classification results are having a higher accuracy with the use of the Nearest Centroid ...
... Also, they have not introduced any feature dimensionality reduction algorithm before the NCC feature classification process. The references [112], and [113] have mentioned, the high dimensional space features provide higher classification accuracies by proving their proposed classification algorithm can achieve over 95% normal and abnormal PCG signal classification accuracy. However, the high dimensional spaced features may have affected by the curse of dimensionality and they may require more time in the classification process which can be avoided by the low dimensional space features. ...
Thesis
An intelligent support system is needed to help the identification of abnormalities of their hearts. The integration of signal processing with feature extraction and different other machine learning techniques, are new research trend in the studies of heart sound analysis. Different features; time domain, frequency domain, and non-linear feature analysis, provide a major impact on heart sound classification. In this research, an automatic heart sound (normal and abnormal) identification method using the first (S1) and second (S2) heart sound features present in Phonocardiography (PCG) signals, has been proposed. The PCG signals are first segmented to identify S1 and S2 prominent peaks, before features from the segmented PCG signals are extracted in both time and frequency domains. The time domain features include prominent peak amplitude, systole and diastole timings, cycle times, area under prominent peaks, and intensity, whilst the frequency domain features are the 20 filter bank energies of Mel Frequency Cepstral Coefficients (MFCC). The selected time and frequency domain features are further expanded using their statistical values of maximum, minimum, average, median, and standard deviation. Supervised Linear Discriminant Analysis (LDA) dimensionality reduction algorithm is applied to the time and frequency domain features, to reduce the number of features and to identify the most reliable features that can help in the classification of the heart sounds. The LDA low dimensional features are classified using the Nearest Centroid Classifier (NCC) and Artificial Neural Network (ANN) classifiers. The normal and abnormal heart sound classification accuracies of the LDA/NCC method are 50%, 33.33%, and 53.33% for time domain, frequency domain, and the combination of both time and frequency domain features, respectively. On the other hand, the proposed LDA/ANN classification method achieves better results than the LDA/NCC method. LDA/ANN with time domain and frequency domain features achieved 90% and 83.33%, respectively, whilst the combined time and frequency domain features achieves 93.33%. Results indicated that the proposed method was able to classify normal and abnormal heart sounds from PCG signals with high accuracy. This indicates that the proposed method may be used by doctors and medical personnel, to assist their diagnosis of patients; by integrating the proposed method with currently available medical devices such as an electronic stethoscope.
... Previous work mostly focused on classifying the PCG signal based on the segmentation approach [11][12][13][14][15][16][17][18][19][20][21]. ...
... Based on PhysioNet2016 datasets [26,27], most researchers proposed segmentation algorithms for obtaining characteristics of heart sounds like the first heart sound (S 1 ) , second heart sound (S 2 ) , and correlated systolic and diastolic periods. Kamson et al. [18] segmented heart sounds based on a modified hidden semi-Markov model (HSMM) with an average F 1 score of 98.38%. Tang et al. [19] derived the multimodal features based on the HSMM segmentation method and predicted the abnormality of heart sounds using the SVM classifier. ...
Article
Full-text available
Diseases associated with the heart are one of the main reasons of death worldwide. Hence, early examination of the heart is important. For analysis of cardiac disorders, a study of heart sounds is a crucial and beneficial approach. Still, automated classification of heart sounds is a challenging task that mainly depends on segmentation of heart sounds and derivation of features using segmented samples. In the literature available for PCG classification provided by PhysioNet/CinC Challenge 2016, most of the research has focused on enhancing the accuracy of the classification model based on complicated segmentation processes and has failed to improve the sensitivity. In this paper, we present an automated heart sound classification by eliminating the segmentation steps using multidomain features, which results in enhanced sensitivity. The study is based on homomorphic envelogram, mel frequency cepstral coefficient (MFCC), power spectral density (PSD), and multidomain feature extraction. The extracted features are trained using the 5-fold cross-validation method based on an ensemble boosting algorithm over 100 independent iterations. Our proposed design is evaluated using public datasets published in PhysioNet/Computers in Cardiology Challenge 2016. Accuracy of 92.47% with improved sensitivity of 94.08% and specificity of 91.95% is achieved using our model. The output performance proves that our proposed model offers superior performance results.
... The detection and analyses of heart sounds are an important and economical means to understand the physiological information of the human body and have proven to be valuable for disease detection and biometric identification [1][2][3][4][5][6][7][8][9]. The most common method for the extraction of heart sound features is time-frequency analysis, represented by the wavelet transform method [4,[10][11][12], and power spectrum analysis, represented by the FFT method [13][14][15][16][17]. Moreover, according to the premise of these analyses, heart-sound signals must be collected in a resting state. ...
... We first introduce a shoulder-strap-type single-channel wireless heart-sound collector and a shoulder-strap-type four-channel wired heart-sound collector and then examine the multiple characteristics of the single-channel heart-sound signal from a resting state to a state of motion. We further analyzed multi-channel heart-sound signals using the graphic representation method, because multi-channel heart-sound signals provide more overall information about the heart than that provided by a single-channel heart-sound signal [3,15,24]. Therefore, based on the idea of integration, we explored a multi-characteristic-representation method for studying the motion of heart sounds. ...
Article
Full-text available
In this paper, the graphic representation method is used to study the multiple characteristics of heart sounds from a resting state to a state of motion based on single- and four-channel heart-sound signals. Based on the concept of integration, we explore the representation method of heart sound and blood pressure during motion. To develop a single- and four-channel heart-sound collector, we propose new concepts such as a sound-direction vector of heart sound, a motion–response curve of heart sound, the difference value, and a state-change-trend diagram. Based on the acoustic principle, the reasons for the differences between multiple-channel heart-sound signals are analyzed. Through a comparative analysis of four-channel motion and resting-heart sounds, from a resting state to a state of motion, the maximum and minimum similarity distances in the corresponding state-change-trend graphs were found to be 0.0038 and 0.0006, respectively. In addition, we provide several characteristic parameters that are both sensitive (such as heart sound amplitude, blood pressure, systolic duration, and diastolic duration) and insensitive (such as sound-direction vector, state-change-trend diagram, and difference value) to motion, thus providing a new technique for the diverse analysis of heart sounds in motion.
... In , the authors proposed to use the hidden semi-Markov model (HSMM) method with logistic regression for segmenting the heart sound signal in a noisy environment. Other proposed methods in recent years include envelop-based methods (Giordano and Knaflitz, 2019;Wei et al., 2019), probabilistic model-based methods (Kamson et al., 2019;Liu et al., 2017;Oliveira et al., 2019) and time/frequency domain based methods (Chen et al., 2017, p. 2;Liu et al., 2018). ...
Preprint
Full-text available
Heart sound diagnosis and classification play an essential role in detecting cardiovascular disorders, especially when the remote diagnosis becomes standard clinical practice. Most of the current work is designed for single category based heard sound classification tasks. To further extend the landscape of the automatic heart sound diagnosis landscape, this work proposes a deep multilabel learning model that can automatically annotate heart sound recordings with labels from different label groups, including murmur's timing, pitch, grading, quality, and shape. Our experiment results show that the proposed method has achieved outstanding performance on the holdout data for the multi-labelling task with sensitivity=0.990, specificity=0.999, F1=0.990 at the segments level, and an overall accuracy=0.969 at the patient's recording level.
... The segmentation of the FHSs is thus an essential step in automatic PCG analysis. The most commonly used heart sounds segmentation methods in recent years include envelope-based methods [9,10], ECG or carotid signal methods [11], probabilistic model methods [12][13][14][15], feature-based methods [16], and timefrequency analysis methods [17]. The utilized algorithms are based on the assumption that the diastolic period is longer than the systolic period. ...
Article
Full-text available
The automated classification of heart sounds plays a significant role in the diagnosis of cardiovascular diseases (CVDs). With the recent introduction of medical big data and artificial intelligence technology, there has been an increased focus on the development of deep learning approaches for heart sound classification. However, despite significant achievements in this field, there are still limitations due to insufficient data, inefficient training, and the unavailability of effective models. With the aim of improving the accuracy of heart sounds classification, an in-depth systematic review and an analysis of existing deep learning methods were performed in the present study, with an emphasis on the convolutional neural network (CNN) and recurrent neural network (RNN) methods developed over the last five years. This paper also discusses the challenges and expected future trends in the application of deep learning to heart sounds classification with the objective of providing an essential reference for further study.
... It is necessary to remove noise prior to processing the recorded sounds. Moreover, the heart of a healthy individual generates different sounds from that of an individual with diseases [3]. A heartbeat generally involves two sounds, i.e., S1 and S2. ...
Preprint
Full-text available
Early diagnosis is crucial in the treatment of heart diseases. Researchers have applied a variety of techniques for cardiovascular disease diagnosis, including the detection of heart sounds. It is an efficient and affordable diagnosis technique. Body organs, including the heart, generate several sounds. These sounds are different in different individuals. A number of methodologies have been recently proposed to detect and diagnose normal/abnormal sounds generated by the heart. The present study proposes a technique on the basis of the Mel-frequency cepstral coefficients, fractal dimension, and hidden Markov model. It uses the fractal dimension to identify sounds S1 and S2. Then, the Mel-frequency cepstral coefficients and the first- and second-order difference Mel-frequency cepstral coefficients are employed to extract the features of the signals. The adaptive Hemming window length is a major advantage of the methodology. The S1-S2 interval determines the adaptive length. Heart sounds are divided into normal and abnormal through the improved hidden Markov model and Baum-Welch and Viterbi algorithms. The proposed framework is evaluated using a number of datasets under various scenarios.
... Other recent improvements to the Viterbi algorithm included searching the sojourn time distribution parameters [26] and a multi-centroid Gaussian model of the diastolic duration [27]. The first improvement employed the expectation maximization algorithm, which was an iterative algorithm that required a large amount of calculations and the calculation time was much longer compared with the original algorithm [3], while it also required long-term signals to achieve better performance, so it was not suitable for real-time segmentation. ...
Article
Full-text available
Heart sound segmentation (HSS) is a critical step in heart sound processing, where it improves the interpretability of heart sound disease classification algorithms. In this study, we aimed to develop a real-time algorithm for HSS by combining the temporal convolutional network (TCN) and the hidden semi-Markov model (HSMM), and improve the performance of HSMM for heart sounds with arrhythmias. We experimented with TCN and determined the best parameters based on spectral features, envelopes, and one-dimensional CNN. However, the TCN results could contradict the natural fixed order of S1-systolic-S2-diastolic of heart sound, and thereby the Viterbi algorithm based on HSMM was connected to correct the order errors. On this basis, we improved the performance of the Viterbi algorithm when detecting heart sounds with cardiac arrhythmias by changing the distribution and weights of the state duration probabilities. The public PhysioNet Computing in Cardiology Challenge 2016 data set was employed to evaluate the performance of the proposed algorithm. The proposed algorithm achieved an F1 score of 97.02%, and this result was comparable with the current state-of-the-art segmentation algorithms. In addition, the proposed enhanced Viterbi algorithm for HSMM corrected 30 out of 30 arrhythmia errors after checking one by one in the dataset.
... These sounds only represent a fraction of each of the CC sections, such that, in stage of most methods relies on the computation of some kind of PCG energy envelope, which is used as a novelty or detection function. Among the techniques that have been used to compute the envelope are, for instance, the Shannon energy [4], wavelets [5][6][7][8][9][10][11][12], Shannon energy combined with the -transform [13], homomorphic filtering [10,11,[14][15][16][17], auto-regressive modeling [18,19] or the Viola integral [20]. In some proposals, the computation of an energy envelope is omitted or replaced by methods such as simplicity measurement [4], atomic time-frequency decomposition using the short-time Fourier transform (STFT) [9,21,22], or empirical mode decomposition followed by a windowed Kurtosis analysis [12,23]. ...
Article
Listening to cardiac sounds can quickly provide information about the functioning of the heart. The heart sound signal, also known as the phonocardiogram (PCG), plays an essential role in automatic auscultation. Segmentation of the PCG signal into its fundamental parts can significantly facilitate any further analysis. In this work, we propose a new method that segments the PCG into fundamental heart sounds and silences. This method can be divided into two stages: Detection and Selection. In the first stage, a function whose maxima indicate the presence of sound events is generated based on the calculation of the spectral flux, a measure of how quickly the spectrum of the PCG signal is changing with respect to time. In the second stage, the position of the beginning and termination of the fundamental heart sounds is detected by analyzing and selectively choosing the time positions of the maxima in the detection function. This selection is solved as an optimization problem through the estimation of an ideal detection function, whose solution is found using two genetic algorithms: a simple genetic algorithm (SGA) and differential evolution (DE). The proposed method was evaluated using the PhysioNet/CinC Challenge dataset, comprising more than 3,000 PCGs. Our results exhibit a mean F1 score of 87.5% and 93.6% for the SGA and DE variants, respectively. The proposed system is robust and highly modular, which simplifies the reuse of specific parts to evaluate algorithm variants. The implementation of the proposed method is available as open-source software.
... They have achieved a comparable F1 score of 95.6% with the state of the art HSS using logistic regression hidden semi-Markov model. [4], [5]. ...
Article
Full-text available
Cardiovascular diseases have become one of the most prevalent threats to human health throughout the world. As a noninvasive assistant diagnostic tool, the heart sound detection techniques play an important role in the prediction of cardiovascular diseases. In this paper, the latest development of the computer-aided heart sound detection techniques over the last five years has been reviewed. There are mainly the following aspects: the theories of heart sounds and the relationship between heart sounds and cardiovascular diseases; the key technologies used in the processing and analysis of heart sound signals, including denoising, segmentation, feature extraction and classification; with emphasis, the applications of deep learning algorithm in heart sound processing. In the end, some areas for future research in computer-aided heart sound detection techniques are explored, hoping to provide reference to the prediction of cardiovascular diseases.
Article
This paper proposes an efficient heart sound segmentation method for automatic detection of heart sounds. In this method, the abrupt change at the heart sound locations is considered as a cue factor for segmentation. The phonocardiogram signal is analysed by passing kurtosis of the signal envelope through zero frequency filter (ZFF). The impulses at the locations of S1 and S2 in the filtered signal are used for the localization of heart sound. The performance of proposed method is evaluated on a real clinical dataset PhysioNet/CinC Challenge Heart Sound (PhysioNet/CinC). A set of 120 heart sound recordings, consisting of normal heart sound as well as pathological heart sound, is considered for evaluation. The experimental result shows that the proposed algorithm achieves an average sensitivity of 98.61%, average positive prediction of 99.11% and average overall accuracy of 98.07%. The robustness of the proposed algorithm is verified using additive white Gaussian noise and respiratory noise.
Chapter
Identifying the precise Heart Sounds (HS) positions inside a Phonocardiogram (PCG); otherwise, Heart Sounds Segmentation (HSS) is a vital phase for the automatic examination of recordings of HS, permitting for the categorization of pathological proceedings. Analysis of HS signals (explicitly, PCG) in the last some decades, particularly for automated HSS and also classification, was largely learned and also stated to encompass the possible value for detecting pathology precisely in medical applications since the bad outcomes in these stages will ruin or shatter the HS detection system’s efficiency. Therefore, the PCG detection issues to implement a new efficient algorithm are required to be discussed. Here, the recently published pre-processing, segmentation, Feature Extractions (FE), and also classification techniques along with their top-notch of PCG signal examination were reviewed. Associated studies are contrasted with their datasets, FE, and the classifiers that they utilized. This effort aims to analyze all the research directions in PCG detection techniques. At the last of this appraisal, several directions for future researches toward PCG signal analysis are rendered.
Chapter
Full-text available
At present, most of the persons are suffering from eyesight. Their food habits and their genes will create eyesight. To test eyesight, there is a manual approach that consumes time as well as money. The noted disadvantages such as the need to take more care personally while interacting with an unknown environment and need to collaborate with other organizations for the charity of testing the eyesights. To avoid these inconveniences, the proposed approach is required which automatically tests the eyesight for right and left eyes using IoT. This smart vision approach using IoT will test eyesight for numerous users and will generate reports separately. That report sent to the user mobile. The virtual software developed uses IoT which enables checking of the sight functionality and the report is sent to the user mobile. In this, the virtual environment is created where IoT devices and networked computer vision devices are connected to test the eyesight. The proposed approach is demanded and is required in this modern and future culture too. Many benefits like consultancy price-cutting transport cost-cutting, time is reduced, speed up in generating reports are achieved. This is a future demanded revolutionary approach which directs many online spectacle shopping sites. This is considered a virtual doctor and serves lots of people since there is no tiredness because it is an automatic approach.
Article
Heart sounds are an important basis for evaluating heart disease. In recent years, auxiliary diagnosis technology of cardiovascular disease based on the detection of heart sound signals has become a research hotspot. A lightweight automatic heart sound classification method is presented in this study. It only needs to perform simple pre-processing on the heart sound data, and a neural network model is constructed to automatically extract the time-frequency features of heart sound data: where the convolution module extracts the frequency domain characteristics, the loop module extracts the time domain characteristics, and the features are then fused in series and parallel. Finally, the classification and recognition of heart sounds are realized based on the fusion features. At the same time, methods such as group convolution, global average pooling, and gated loop mechanisms are used to reduce the parameters and training time of the neural network. The experimental data from the PhysioNet database is used for testing. The recognition accuracy of the series feature fusion heart sound classification model is 95.00%, and the model size is 0.6M. Additionally, the parallel feature fusion heart sound classification model is 95.50%, and the model size is 1.36M. Therefore, the lightweight heart sound classification model designed in this study contains fewer parameters and higher recognition accuracy. This makes it suitable for deployment in embedded devices. It also has important research significance for the development of portable heart sound detection equipment.
Article
This paper presents a new method to extract the envelope of the fundamental heart sound (S1 and S2) using the logistic function. The sigmoid characteristic of the logistic function is incorporated to segregate S1, and S2 signal intensities from silent or noise interfered systolic and diastolic intervals in a heart sound cycle. This signal intensity transformation brings uniformity to the envelope peak of S1 and S2 sound by inclining the transform intensity distribution towards the upper asymptote of the sigmoid curve. The proposed logistic function based amplitude moderation (LFAM) envelogram method involves finding the critical upper amplitude (xuc) above which the signals will be categorized as loud sound and the critical lower amplitude (xlc) below which the signal will be considered as noise. These critical values are regressively obtained from the signal itself by histogram analysis of intensity distribution. The performance is evaluated on noisy PCG dataset taken from PhysioNet/Computing in Cardiology Challenge 2016. The LFAM envelope yields better hill-valley discrimination of heart sounds from its silent/noisy signal intervals. The enhance heart sound envelope peaks are better than conventional methods. The proposed envelope feature is evaluated for heart sound segmentation using HSMM. There is a significant improvement in segmentation accuracy, especially at a low signal-to-noise ratio. The best average F1 score is 97.73%.
Conference Paper
Full-text available
Auscultation is the primary tool for detection and diagnosis of cardiovascular diseases in hospitals and home visits. This fact has led in the recent years to the development of automatic methods for heart sound classification, thus allowing for detecting cardiovascular pathologies in an effective way. The aim of this paper is to review recent methods for automatic classification and to apply several signal processing techniques in order to evaluate them in the PhysioNet/CinC Challenge 2016 results. For this purpose, the records of the open database PysioNet/Computing are modified by segmentation or filtering methods and the results were tested using the challenge best ranked algorithms. Results show that an adequate preprocessing of data and subsequent feature selection may improve the performance of machine learning and classification techniques.
Conference Paper
Full-text available
In the past few decades heart sound signals (i.e., phono-cardiograms or PCGs) have been widely studied. Automated heart sound segmentation and classification techniques have the potential to screen for pathologies in a variety of clinical applications. However, comparative analyses of algorithms in the literature have been hindered by the lack of a large and open database of heart sound recordings. The PhysioNet/Computing in Cardiology (CinC) Challenge 2016 addresses this issue by assembling the largest public heart sound database, aggregated from eight sources obtained by seven independent research groups around the world. The database includes 4,430 recordings taken from 1,072 subjects, totalling 233,512 heart sounds collected from both healthy subjects and patients with a variety of conditions such as heart valve disease and coronary artery disease. These recordings were collected using heterogeneous equipment in both clinical and nonclinical (such as in-home visits). The length of recording varied from several seconds to several minutes. Additional data provided include subject demographics (age and gender), recording information (number per patient, body location, and length of recording), synchronously recorded signals (such as ECG), sampling frequency and sensor type used. Participants were asked to classify recordings as normal, abnormal, or not possible to evaluate (noisy/uncertain). The overall score for an entry was based on a weighted sensitivity and specificity score with respect to manual expert annotations. A brief description of a baseline classification method is provided, including a description of open source code, which has been provided in association with the Challenge. The open source code provided a score of 0.71 (Se=0.65 Sp=0.76). During the official phase of the competition, a total of 48 teams submitted 348 open source entries, with a highest score of 0.86 (Se=0.94 Sp=0.78).
Article
Full-text available
In the past few decades, analysis of heart sound signals (i.e. the phonocardiogram or PCG), especially for automated heart sound segmentation and classification, has been widely studied and has been reported to have the potential value to detect pathology accurately in clinical applications. However, comparative analyses of algorithms in the literature have been hindered by the lack of high-quality, rigorously validated, and standardized open databases of heart sound recordings. This paper describes a public heart sound database, assembled for an international competition, the PhysioNet/Computing in Cardiology (CinC) Challenge 2016. The archive comprises nine different heart sound databases sourced from multiple research groups around the world. It includes 2435 heart sound recordings in total collected from 1297 healthy subjects and patients with a variety of conditions, including heart valve disease and coronary artery disease. The recordings were collected from a variety of clinical or nonclinical (such as in-home visits) environments and equipment. The length of recording varied from several seconds to several minutes. This article reports detailed information about the subjects/patients including demographics (number, age, gender), recordings (number, location, state and time length), associated synchronously recorded signals, sampling frequency and sensor type used. We also provide a brief summary of the commonly used heart sound segmentation and classification methods, including open source code provided concurrently for the Challenge. A description of the PhysioNet/CinC Challenge 2016, including the main aims, the training and test sets, the hand corrected annotations for different heart sound states, the scoring mechanism, and associated open source code are provided. In addition, several potential benefits from the public heart sound database are discussed.
Article
Full-text available
An efficient heart sound segmentation (HSS) method that automatically detects the location of first ( S1) and second ( S2) heart sound and extracts them from heart auscultatory raw data is presented here. The heart phonocardiogram is analyzed by employing ensemble empirical mode decomposition (EEMD) combined with kurtosis features to locate the presence of S1, S2, and extract them from the recorded data, forming the proposed HSS scheme, namely HSS-EEMD/K. Its performance is evaluated on an experimental dataset of 43 heart sound recordings performed in a real clinical environment, drawn from 11 normal subjects, 16 patients with aortic stenosis, and 16 ones with mitral regurgitation of different degrees of severity, producing 2608 S1 and S2 sequences without and with murmurs, respectively. Experimental results have shown that, overall, the HSS-EEMD/K approach determines the heart sound locations in a percentage of 94.56% and segments heart cycles correctly for the 83.05% of the cases. Moreover, results from a noise stress test with additive Gaussian noise and respiratory noises justify the noise robustness of the HSS-EEMD/K. When compared with four other efficient methods that mainly employ wavelet transform, energy, simplicity, and frequency measures, respectively, using the same experimental database, the HSS-EEMD/K scheme exhibits increased accuracy and prediction power over all others at the level of 7-19% and 4-9%, respectively, both in controls and pathological cases. The promising performance of the HSS-EEMD/K paves the way for further exploitation of the diagnostic value of heart sounds in everyday clinical practice.
Article
Full-text available
In automated heart sound analysis and diagnosis, a set of clinically valued parameters including sound intensity, frequency content, timing, duration, shape, systolic and diastolic intervals, the ratio of the first heart sound amplitude to second heart sound amplitude (S1/S2), and the ratio of diastolic to systolic duration (D/S) is measured from the PCG signal. The quality of the clinical feature parameters highly rely on accurate determination of boundaries of the acoustic events (heart sounds S1, S2, S3, S4 and murmurs) and the systolic/diastolic pause period in the PCG signal. Therefore, in this paper, we propose a new automated robust heart sound activity detection (HSAD) method based on the total variation filtering, Shannon entropy envelope computation, instantaneous phase based boundary determination, and boundary location adjustment. The proposed HSAD method is validated using different clean and noisy pathological and non-pathological PCG signals. Experiments on a large PCG database show that the HSAD method achieves an average sensitivity (Se) of 99.43% and positive predictivity (+P) of 93.56%. The HSAD method accurately determines boundaries of major acoustic events of the PCG signal with signal-to-noise ratio of 5 dB. Unlike other existing methods, the proposed HSAD method does not use any search-back algorithms. The proposed HSAD method is a quite straightforward and thus it is suitable for real-time wireless cardiac health monitoring and electronic stethoscope devices.
Article
Full-text available
In many examples in biomedical signal processing the modulation amplitude of a signal (e.g. respira-tion movement) conveys information. Traditional methods of demodulation typically involve peak-and valley tracing which is computationally expensive and noise-sensitive. In this paper we present results for the well-known, but little-used, technique of complex-domain homomorphic ltering which ooers an alternative approach to such demodulation problems.
Conference Paper
Full-text available
This work presents a novel method for automatic detection and identification of heart sounds. Homomorphic filtering is used to obtain a smooth envelogram of the phono cardiogram, which enables a robust detection of events of interest in heart sound signal. Sequences of features extracted from the detected events are used as observations of a hidden Markov model. It is demonstrated that the task of detection and identification of the major heart sounds can be learned from unlabelled phono cardiograms by an unsupervised training process and without the assistance of any additional synchronizing channels
Conference Paper
Full-text available
The monitoring of respiration rates using impedance plethysmography is often confused by cardiac activity. This paper proposes using the phonocardiogram as an alternative, since the process of respiration affects heart sounds. As part of this research, a technique is developed to segment heart sounds into its component segments, using hidden Markov models. The heart sounds data is preprocessed into feature vectors, where the feature vectors are comprised of the average Shannon energy of the heart sound signal, the delta Shannon energy, and the delta-delta Shannon energy. The performance of the segmentation system is validated using eight-fold cross-validation
Article
Full-text available
Total variation (TV) is a powerful method that brings great benefit for edge-preserving regularization. Despite being widelyemployed inimage processing,ithas restricted applicability for 1-D signal processing since piecewise-constant signals form a rather limited model for many applications. Here we generalize conventional TV in 1-D by extending the derivative operator, which is within the regularization term, to any linear differential operator. This provides flexibility for tailoring the approach to the presence of nontrivial linear systems and for different types of driving signals such as spike-like, piecewise-constant, and so on. Conventional TV remains a special case of this general frame- work. We illustrate the feasibility of the method by considering a nontrivial linear system and different types of driving signals.
Article
Full-text available
To develop a digital algorithm that detects first and second heart sounds, defines the systole and diastole, and characterises the systolic murmur. Heart sounds were recorded in 300 children with a cardiac murmur, using an electronic stethoscope. A Digital algorithm was developed for detection of first and second heart sounds. R-waves and T-waves in the electrocardiography were used as references for detection. The sound signal analysis was carried out using the short-time Fourier transform. The first heart sound detection rate, with reference to the R-wave, was 100% within 0.05-0.2R-R interval. The second heart sound detection rate between the end of the T-wave and the 0.6R-R interval was 97%. The systolic and diastolic phases of the cardiac cycle could be identified. Because of the overlap between heart sounds and murmur a systolic segment between the first and second heart sounds (20-70%) was selected for murmur analysis. The maximum intensity of the systolic murmur, its average frequency, and the mean spectral power were quantified. The frequency at the point with the highest sound intensity in the spectrum and its time from the first heart sound, the highest frequency, and frequency range were also determined. This method will serve as the foundation for computer-based detection of heart sounds and the characterisation of cardiac murmurs.
Article
Full-text available
Heart murmurs are often the first signs of pathological changes of the heart valves, and they are usually found during auscultation in the primary health care. Distinguishing a pathological murmur from a physiological murmur is however difficult, why an "intelligent stethoscope" with decision support abilities would be of great value. Phonocardiographic signals were acquired from 36 patients with aortic valve stenosis, mitral insufficiency or physiological murmurs, and the data were analyzed with the aim to find a suitable feature subset for automatic classification of heart murmurs. Techniques such as Shannon energy, wavelets, fractal dimensions and recurrence quantification analysis were used to extract 207 features. 157 of these features have not previously been used in heart murmur classification. A multi-domain subset consisting of 14, both old and new, features was derived using Pudil's sequential floating forward selection (SFFS) method. This subset was compared with several single domain feature sets. Using neural network classification, the selected multi-domain subset gave the best results; 86% correct classifications compared to 68% for the first runner-up. In conclusion, the derived feature set was superior to the comparative sets, and seems rather robust to noisy data.
Article
Full-text available
Motivated by the classical TV (total variation) restoration model, we propose a new nonlinear filter-the digital TV filter for denoising and enhancing digital images, or more generally, data living on graphs. The digital TV filter is a data dependent lowpass filter, capable of denoising data without blurring jumps or edges. In iterations, it solves a global total variational (or L(1)) optimization problem, which differs from most statistical filters. Applications are given in the denoising of one dimensional (1-D) signals, two-dimensional (2-D) data with irregular structures, gray scale and color images, and nonflat image features such as chromaticity.
Article
Full-text available
We have previously developed a method for localization of the first heart sound (S1) using wavelet denoising and ECG-gated peak-picking. In this study, an additional enhancement step based on cross-correlation and ECG-gated ensemble averaging (EA) is presented. The main objective of the improved method was to localize S1 with very high temporal accuracy in (pseudo-) real time. The performance of S1 detection and localization, with and without EA enhancement, was evaluated on simulated as well as experimental data. The simulation study showed that EA enhancement reduced the localization error considerably and that S1 could be accurately localized at much lower signal-to-noise ratios. The experimental data were taken from ten healthy subjects at rest and during invoked hyper- and hypotension. For this material, the number of correct S1 detections increased from 91% to 98% when using EA enhancement. Improved performance was also demonstrated when EA enhancement was used for continuous tracking of blood pressure changes and for respiration monitoring via the electromechanical activation time. These are two typical applications where accurate localization of S1 is essential for the results.
Conference Paper
Full-text available
Desribes the development of a segmentation algorithm which separates the heart sound signal into four parts: the first heart sound, the systole, the second heart sound and the diastole. The segmentation of phonocardiogram (PCG) signals is the first step of analysis and the most important procedure in the automatic diagnosis of heart sounds. This algorithm is based on the normalized average Shannon energy of a PCG signal. The performance of the algorithm has been evaluated using 515 periods of PCG signals recording from 37 objects including normal and abnormal. The algorithm has achieved a 93 percent correct ratio
Article
This study concerns the task of automatic structural heart abnormality risk detection from digital phonocardiogram (PCG) signals aiming at pediatric heart disease screening applications. Recently, various systems based on convolutional neural networks trained on time-frequency representations of segmental PCG frames have been presented that outperform systems using hand-crafted features. This study focuses on the segmentation and time-frequency representation components of the CNN-based designs. We consider the most commonly used features (MFCC and Mel-Spectrogram) used in state-of-the-art systems and a time-frequency representation influenced by domain-knowledge, namely sub-band envelopes as an alternative feature. Via tests carried on two high quality databases with a large set of possible settings, we show that sub-band envelopes are preferable to the most commonly used features and period synchronous windowing is preferable over asynchronous windowing.
Article
Phonocardiogram (PCG) signal analysis is an effective and convenient method for the preliminary diagnosis of heart disease. In this study, a scaled spectrogram and partial least squares regression (PLSR) based method was proposed for the classification of PCG signals. Proposed method is mainly comprised of four stages, namely as being heart cycle estimation, spectrogram scaling, dimension reduction and classification. At the heart cycle estimation stage, the short time average magnitude difference of the Shannon energy envelope is applied. Then the spectrogram of the obtained heart cycle is calculated for feature extraction. However, the sizes of the spectrograms between different PCG signals are usually not the same. In order to overcome the difficulty of direct comparison, the bilinear interpolation is used for the spectrogram to get the scaled spectrogram with a fixed size. Nevertheless, the scaled spectrogram contains a large quantity of redundant and irrelevant information. To extract the most relevant features from the scaled spectrogram, we adopt the PLSR to reduce the dimension of the scaled spectrograms. Since PLSR has the advantage of using the category information during the dimension reduction process, the extracted features are more discriminative. Then the classification results are obtained via support vector machine (SVM). The proposed method is evaluated on two public datasets offered by the PASCAL classifying heart sounds challenge, and the results are compared to those obtained using the best methods in the challenge, thereby proving the effectiveness of our method.
Conference Paper
In this paper an automatic heart sound segmentation scheme is proposed using Teager-Kaiser energy operator (TKEO) in Hilbert space. After pre-processing the heart sound signal, Hilbert envelope is computed and filtering is performed to obtain smooth envelope. TKEO is applied to this envelope signal to find more accurate boundaries of heart sounds. This energy operator depends only on first two time derivatives of the signal. It gives suitable boundaries for first and second heart sound separation. The heart sound boundaries for S1 and S2 are decided by markers derived from second derivative of Teager energy envelope. The heart sound signals are downloaded from the web site of the Department of Medicine, Washington University to evaluate the proposed algorithm. The segmented heart sound components along-with original recorded signals are heard by a number of signal processing experts. They found the performance of the proposed method satisfactory.
Article
The identification of the exact positions of the first and second heart sounds within a phonocardiogram (PCG), or heart sound segmentation, is an essential step in the automatic analysis of heart sound recordings, allowing for the classification of pathological events. While threshold-based segmentation methods have shown modest success, probabilistic models, such as hidden Markov models, have recently been shown to surpass the capabilities of previous methods. Segmentation performance is further improved when a priori information about the expected duration of the states is incorporated into the model, such as in a hidden semi-Markov model (HSMM). This article addresses the problem of the accurate segmentation of the first and second heart sound within noisy, real-world PCG recordings using a HSMM, extended with the use of logistic regression for emission probability estimation. In addition, we implement a modified Viterbi algorithm for decoding the most-likely sequence of states, and evaluated this method on a large dataset of 10172 seconds of PCG recorded from 112 patients (including 12181 first and 11627 second heart sounds). The proposed method achieved an average F1 score of 95.630.85% while the current state-ofthe- art achieved 86.281.55% when evaluated on unseen test recordings. The greater discrimination between states afforded using logistic regression as opposed to the previous Gaussian distribution-based emission probability estimation as well as the use of an extended Viterbi algorithm allows this method to significantly outperform the current state-of-the-art method based on a two-sided, paired t-test.
Article
The Ward error sum of squares hierarchical clustering method has been very widely used since its first description by Ward in a 1963 publication. It has also been generalized in various ways. Two algorithms are found in the literature and software, both announcing that they implement the Ward clustering method. When applied to the same distance matrix, they produce different results. One algorithm preserves Ward’s criterion, the other does not. Our survey work and case studies will be useful for all those involved in developing software for data analysis using Ward’s hierarchical clustering method.
Conference Paper
In this study, a methodology based on hidden Markov models (HMM) as a probabilistic finite state-machine to model systolic and diastolic interval duration is proposed. The detection of the first (S1) and second (S2) heart sound is performed using a network of two HMM's with grammar constraints to parse sequence of systolic and diastolic intervals. Duration modeling was considered in the HMM model architecture selection based on experimental measurements of systolic and diastolic intervals in normal subjects. Feature extraction of heart sound signals was based on time-cepstral features. Results are presented in terms of detection performance compared with QRS peak annotations of the simultaneous ECG recording. The performance of the proposed approach has been evaluated in 80 subjects. The results showed that the system was effective to detect the first and second heart sounds with sensitivity of 95% and a positive predictive value of 97% and thus provides a promising methodology for heart sound analysis.
Article
A novel method for segmentation of heart sounds (HSs) into single cardiac cycle (S1-Systole-S2-Diastole) using homomorphic filtering and K-means clustering is presented. Feature vectors were formed after segmentation by using Daubechies-2 wavelet detail coefficients at the second decomposition level. These feature vectors were then used as input to the neural networks. Grow and Learn (GAL) and Multilayer perceptron-Backpropagation (MLP-BP) neural networks were used for classification of three different HSs (Normal, Systolic murmur and Diastolic murmur). It was observed that the classification performance of GAL was similar to MLP-BP. However, the training and testing times of GAL were lower as compared to MLP-BP. The proposed framework could be a potential solution for automatic analysis of HSs that may be implemented in real time for classification of HSs.
Article
Phonocardiograms (PCGs), recordings of heart sounds, have many advantages over traditional auscultation in that they may be replayed and analysed for spectral and frequency information. PCG is not a widely used diagnostic tool as it could be. One of the major problems with PCG is noise corruption. Many sources of noise may pollute a PCG including foetal breath sounds if the subject is pregnant, lung and breath sounds, environmental noise and noise from contact between the recording device and the skin. An electronic stethoscope is used to record heart sounds and the problem of extracting noise from the signal is addressed via the use of wavelets and averaging. Using the discrete wavelet transform, the signal is decomposed. Due to the efficient decomposition of heart signals, their wavelet coefficients tend to be much larger than those due to noise. Thus, coefficients below a certain level are regarded as noise and are thresholded out. The signal can then be reconstructed without significant loss of information in the signal content. The questions that this study attempts to answer are which wavelet families, levels of decomposition, and thresholding techniques best remove the noise in a PCG. The use of averaging in combination with wavelet denoising is also addressed. Possible applications of the Hilbert transform to heart sound analysis are discussed.
Article
We survey agglomerative hierarchical clustering algorithms and discuss efficient implementations that are available in R and other software environments. We look at hierarchical self-organizing maps, and mixture models. We review grid-based clustering, focusing on hierarchical density-based approaches. Finally, we describe a recently developed very efficient (linear time) hierarchical clustering algorithm, which can also be viewed as a hierarchical grid-based algorithm. © 2011 Wiley Periodicals, Inc.
Article
In this study, the valvular heart disorder (VHD) detection method by the wavelet packet (WP) decomposition and the support vector machine (SVM) techniques are proposed. From considering the truth that the frequency ranges of the normal sound and VHDs are different from each other, the WP decomposition at level 8 is employed to split more elaborate frequency bandwidths of the heart sound signals. And then the WP energy (WPE) with the distribution information of energy throughout the whole frequency range of heart sound signals is calculated. Since the heart sound signals with the frequency range of 20–750 Hz are preferred in this study, WPEs at the terminal nodes from (8, 1) to (8, 47) are selected and two parameters meanWPE and stdWPE as defined by the mean value and standard deviation of the position indices of the terminal nodes with over the weighting value (ζ) of the maximum value of WPE are proposed as a feature. Furthermore, the SVM technique is employed as the identification tool to classify between the normal sound and VHDs. Finally, a case study on the normal sound, aortic and mitral VHDs is demonstrated to validate the usefulness and efficiency of the VHD detection using WP decomposition and SVM classifier. The experimental results of the proposed VHD detection method showed high performance like the specificity of over 96% and the sensitivity of 100% for both the training and testing data.
Article
A sensitive technic for frequency analysis of the first heart sound (S1) during isovolumic contraction time (ICT) was developed. Recorded heart sounds were filtered with a dynamic analyzer. Amplitudes of frequencies between 30 and 70 cps were plotted as a percentage of peak total energy of S1 against frequency. A consistent and reproducible frequency "fingerprint" was obtained in 74 normal subjects. Frequencies of S1 were shown to be directly proportional to ventricular elasticity (VE) and inversely proportional to combined ventricular mass (VM). VM is constant during ICT. Amplitude at 40 cps was less than at 30 cps because of reduced VE (myocardial infarction), increased VM (athletes), or combined reduction in VE and increased VM (myocardiopathy). Normal patterns were found in aortic insufficiency (increased VE and VM). Diagnostic patterns were found in 21 of 24 patients with acute myocardial infarction (MI) but were similar to patterns found in patients with healed infarcts and myocardiopathy and athletes. Acute pulmonary embolism could be differentiated from MI.
Article
Digital stethoscopes offer new opportunities for computerized analysis of heart sounds. Segmentation of heart sound recordings into periods related to the first and second heart sound (S1 and S2) is fundamental in the analysis process. However, segmentation of heart sounds recorded with handheld stethoscopes in clinical environments is often complicated by background noise. A duration-dependent hidden Markov model (DHMM) is proposed for robust segmentation of heart sounds. The DHMM identifies the most likely sequence of physiological heart sounds, based on duration of the events, the amplitude of the signal envelope and a predefined model structure. The DHMM model was developed and tested with heart sounds recorded bedside with a commercially available handheld stethoscope from a population of patients referred for coronary arterioangiography. The DHMM identified 890 S1 and S2 sounds out of 901 which corresponds to 98.8% (CI: 97.8-99.3%) sensitivity in 73 test patients and 13 misplaced sounds out of 903 identified sounds which corresponds to 98.6% (CI: 97.6-99.1%) positive predictivity. These results indicate that the DHMM is an appropriate model of the heart cycle and suitable for segmentation of clinically recorded heart sounds.
Article
Heart sound is a valuable biosignal for early detection of a large set of cardiac diseases. Ambient and physiological noise interference is one of the most usual and high probable incidents during heart sound acquisition. It may change the prominent and crucial characteristics of heart sound which may possess important information for heart disease diagnosis. In this paper, we propose a new method to detect ambient and internal body noises in heart sounds. The algorithm utilizes physiologically inspired periodicity/semi-periodicity criteria. A small segment of clean heart sound exhibiting periodicity in the time and in the frequency domain is first detected. The sound segment is used as a template to detect uncontaminated heart sounds during recording. The technique has been tested on the heart sounds contaminated with several types of noises, recorded from 68 different subjects. Average sensitivity of 95.13% and specificity of 98.65% for non-cardiac sound detection were achieved.
Article
Before heart rate (HR) variability can be used for predictive purposes in the clinical setting, day-to-day variation and reproducibility need to be defined as do relations to mean HR. HR variability and mean HR were therefore determined in 2 successive 24-hour ambulatory electrocardiograms obtained from 33 normal subjects (age 34 +/- 7 years, group I), and 22 patients with coronary disease and stable congestive heart failure (CHF) (age 59 +/- 7 years, group II). Three measures were used: (1) SDANN (standard deviation of all mean 5-minute normal sinus RR intervals in successive 5-minute recording periods over 24 hours); (2) SD (the mean of the standard deviation of all normal sinus RR intervals in successive 5-minute recording periods over 24 hours); and (3) CV (coefficient of variation of the SD measure), a new measure that compensates for HR effects. Group mean HR was higher and HR variability lower in group II than in group I (80 +/- 10 vs 74 +/- 9 beats/min, p less than 0.04). Mean group values for HR and HR variability showed good correlations between days 1 and 2 (mean RR, r = 0.89, 0.97; SDANN, r = 0.87, 0.87; SD, r = 0.93, 0.97; CV, r = 0.95, 0.97 in groups I and II, respectively). In contrast, considerable individual day-to-day variation occurred (group I, 0 to 46%; group II, 0 to 51%). Low HR variability values were more consistent than high values. SDANN and SD correlated moderately with HR in both groups (r = 0.50 to 0.64). The CV measure minimizes HR effects on HR variability.(ABSTRACT TRUNCATED AT 250 WORDS)
Article
Many disease of the heart cause changes in heart sounds and additional murmurs before other signs and symptoms appear. Hence, heart sound analysis by auscultation is the primary test conducted by physicians to assess the condition of the heart. Yet, heart sound analysis by auscultation as well as analysis of the phonocardiogram (PCG) signal have not gained widespread acceptance. This is due mainly to many controversies regarding the genesis of the sounds and the lack of quantitative techniques for reliable analysis of the signal features. The heart sound signal has much more information than can be assessed by the human ear or by visual inspection of the signal tracings on paper as currently practiced. Here, we review the nature of the heart sound signal and the various signal-processing techniques that have been applied to PCG analysis. Some new research directions are also outlined.
Article
The diagnosis of autonomic neuropathy frequently depends on results of tests which elicit reflex changes in heart rate. Few well-documented normal ranges are available for these tests. The present study was designed to investigate the effect of age upon heart rate variability at rest and in response to a single deep breath, the Valsalva manoeuvre, and standing. A computerised method of measurement of R-R interval variation was used to study heart rate responses in 310 healthy subjects aged 18-85 years. Heart rate variation during each procedure showed a skewed distribution and a statistically significant negative correlation with age. Normal ranges (90% and 95% confidence limits) for subjects aged 20-75 years were calculated for heart rate difference (max-min) and ratio (max/min) and standard deviation (SD). Heart rate responses were less than the 95th centile in at least one of the four procedures in 39 (12.6%) out of the 310 subjects, and were below this limit in two or more tests in five (1.6%) subjects. In view of the decline in heart rate variation with increasing age, normal ranges for tests of autonomic function must be related to the age of the subject.
Article
Frequency analysis of heart sounds has been gaining recognition as a possible indicator of several heart and valve diseases, although a comprehensive study of normal heart sounds has not been published. Relating the frequency content of normal heart sounds to certain physical characteristics surrounding the generation of these sounds could lead to a valuable diagnostic tool and give a better understanding of the mechanism of heart sounds production. In this study, the first and second heart sounds from seventy-four normal, and seven hypertensive volunteers were recorded, digitized and analysed using a Fast Fourier Transform algorithm. Statistical analysis was used to relate physical characteristics (sex, blood pressure, and body surface area) of the subjects to the frequency content of normal heart sounds and to compare normal and hypertensive heart sounds. Statistical analysis showed that the major concentration of energy, for both first heart sound (S1) and second heart sound (S2), is below 150 Hertz (Hz) which may indicate that both sounds are caused by vibrations within the same structure, possibly the entire heart. However S2 spectra have greater amplitude than S1 spectra above 150 Hz, which may be due to vibrations within the aorta and pulmonary artery. Relationships observed between body surface area, sex, blood pressure, and the frequency content of heart sounds indicate that as heart size increases, the amplitude of the frequency coefficients above 150 Hz decreases. These observations were more identifiable in the S1 spectra than in the S2 spectra, possibly because the S2 higher frequency components may mask subtle changes in the S2 spectra caused by heart size changes. However, when the changes in heart size are significant, as in hypertension or increased body surface area, trends in the S2 spectra can be observed.
Article
We present here details of a microcomputer system developed to segment the phonocardiogram signal (PCG) and characterize murmurs. Using the ECG and carotid pulse as references, the PCG is segmented into systolic and diastolic parts. Four parameters representing the time and frequency domain characteristics of the signal segments are then computed. Results of application of the methods to 47 phonocardiogram signals are presented. The use of these parameters for the detection and classification of murmurs is discussed.
Total variation filtering, White paper
  • I W Selesnick
  • I Bayram
I.W. Selesnick, I. Bayram, Total variation filtering, White paper, 2010.