A framework for automatic heart sound analysis without segmentation

Abstract

BackgroundA 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.MethodEqual 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.ResultThe 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.ConclusionThe 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.

Survey of Different Methodologies Used in Phonocardiogram Signal Analysis

“…Using this algorithm there was no need to label the individual Fundamental Heart Sounds (FHS) for extraction of individual heart cycles. Discrete wavelet transform was used for feature extraction of cardiac sounds and classification was done using neural network bragging predictors [7]. In 2011, Sanei et al used Adaptive Singular Spectrum Analysis (ASSA) to separate murmurs from the recorded heart sounds.…”

Diagnosis of Heart Valve Disorders through Trapezoidal Features and Hybrid Classifier

“…Babaei et al [17] utilized the main statistical characteristics of PCG signals such as mean and standard deviation in different level of DWT decomposition and acquired 94.24% accuracy for classifying AR, AS and PS (pulmonary stenosis). A framework was proposed by Yuenyong et al [6] based on DWT and without segmentation, for heart sound classification and gained 92% accuracy to classify a number of different heart valve disorders. Studies were reported by Choi et al [2], [10] based on WPT for classifying heart sounds.…”

“…identification of the first and the second heart sound (S1 & S2). There are a number of researches that were addressed PCG segmentation [4], [5], although some works were reported for classifying heart sound without segmentation [6]- [8]. In order to extract discriminant features, an appropriate signal analysis technique is required.…”

Heart Sound Anomaly and Quality Detection using Ensemble of Neural Networks without Segmentation

“…Some segmentation approaches use envelope extraction based on wavelet transform to gain the frequency characteristics of S1 and S2 sound [6]. In the second type approaches, abnormal PCG records are detected without segmentation [7,8].…”