Detection of S1 and S2 Heart Sounds by High Frequency Signatures

Centre for Informatics and Systems, University of Coimbra, Portugal.
Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 02/2006; 1:1410-6. DOI: 10.1109/IEMBS.2006.260735
Source: PubMed


A new unsupervised and low complexity method for detection of S1 and S2 components of heart sound without the ECG reference is described The most reliable and invariant feature applied in current state-of-the-art of unsupervised heart sound segmentation algorithms is implicitly or explicitly the S1-S2 interval regularity. However; this criterion is inherently prone to noise influence and does not appropriately tackle the heart sound segmentation of arrhythmic cases. A solution based upon a high frequency marker; which is extracted from heart sound using the fast wavelet decomposition, is proposed in order to estimate instantaneous heart rate. This marker is physiologically motivated by the accentuated pressure differences found across heart valves, both in native and prosthetic valves, which leads to distinct high frequency signatures of the valve closing sounds. The algorithm has been validated with heart sound samples collected from patients with mechanical and bio prosthetic heart valve implants in different locations, as well as with patients with native valves. This approach exhibits high sensitivity and specificity without being dependent on the valve type nor their implant position. Further more, it exhibits invariance with respect to normal sinus rhythm (NSR) arrhythmias and sound recording location.

Full-text preview

Available from:
  • Source
    • "Diagram showing the movement of blood through the heart A key challenge in automated cardiac auscultation is heart sound segmentation, which refers to the identification of the physiological features of the different heart sounds to determine their relative physiological timing. Several techniques have been developed to detect the timing of the sounds, using peaks in the energy envelope [2], high frequency signatures in wavelet coefficients [3], neural networks [4], empirical mode decomposition [5], Hidden Markov Models [6] and selectional regional correlation [7]. B. Contribution The ideal Heart Sound Segmentation system should be able to accurately identify S1 and S2, use as few assumptions about the signal as possible to reduce the effect of noise, should avoid the use of prior learning to increase generality across patients and conditions, and should be able to work on a small segment of a signal so that it can be used in a close to real time environment. "

    Full-text · Conference Paper · Oct 2015
  • Source
    • "In order to detect the heart cycles, an adaptive threshold is defined for this envelope. The algorithm aims to detect the high frequency signatures (HFS) and the low frequency signatures (LFS) [18]. Kumar et al. consider that usually S2 sounds contain higher frequency with respect to S1 sound (HFS correspond to S2 and LFS correspond to S1) excluding some rare excep- tions. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper considers the problem of classification of the first and the second heart sounds (S1 and S2) under cardiac stress test. The main objective is to classify these sounds without electrocadiogram (ECG) reference and without taking into consideration the systolic and the diastolic time intervals criterion which can become problematic and useless in several real life settings as severe tachycardia, tachyarrhythmia or in the case of subjects under cardiac stress activity. First, the heart sounds are segmented by using a modified time-frequency based envelope. Then to distinguish between the first and the second heart sounds, new features based on high order statistics and energy concentration measures of the Stockwell Transform (S-Transform) are proposed in this study. The first feature named αopt corresponds to the optimal width of the Gaussian window that maximizes the energy concentration of the signal. The second feature named β is the integration over time of the envelope obtained by a modified measure of the instantaneous frequency of the signal. The third feature named ɤ calculates the kurtosis of the time-frequency envelope. A study of the variation of the high frequency content of S1 and S2 over the HR (heart rate) is also discussed. The proposed features are validated on a database that contains 2636 S1 and S2 sounds (1318 S1 and 1318 S2) corresponding to 62 heart signals and 8 subjects under cardiac stress test collected from healthy subjects. Results and comparisons with existing methods in the literature show a large superiority for our proposed features; AUC=0.96 for the proposed ɤ feature and 0.6 and for the existing feature in the literature.
    Full-text · Article · Feb 2015 · Computational and Mathematical Methods in Medicine
  • Source
    • "The segmentation method based on the signal's envelop is basically formed by two simple steps [13]: (i) first the S1 and S2 candidates are identified using the zero-crossings of the envelop of the approximation coefficients of the 5 th level wavelet decomposition. The envelop is computed with a running average of the Shannon energy. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We present a Matlab framework for heart sound processing and analysis. This framework includes algorithms developed for segmentation of the main heart sound components capable of handling situations with high-grade murmur and systolic time intervals (STI) measurement using heart sound. Methods for cardiac function parameter extraction based on STI are also included. Currently, the proposed algorithms are being extended for multi-channel applications. Most of the algorithms outlined in the paper have been extensively evaluated using data collected from patients with several types of cardiovascular diseases under real-life conditions.
    Full-text · Conference Paper · Jan 2014
Show more