English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Segmenting heart sounds accurately is significant to determine the types and the severity of heart diseases. Using cyclostationary property of heart sounds and clustering method, the paper proposed an effective automatic segmentation algorithm of heart sound signals. This algorithm can segment normal heart sounds and abnormal heart sounds with medium murmur exactly. Simulation results show that this method is robust to noise, which is helpful to feature extraction, heart sound recognition and computer-aided diagnosis afterwards. Introduction Phonocardiography is a noninvasive, easy, and efficient method to evaluate heart functions. Heart auscultation is widely used in medical check-up for adults and children due to its easy operation and economy. In auscultation, physicians or cardiologists try to separate heart sound signal into four phases, i.e., first heart sounds (S1), systolic phase, second heart sounds (S2), diastolic phase, and then analyze the features. However, heart sounds are transient and fast varying. Heart sounds analysis by human ears highly depends on the skills and experience of the listeners. Therefore, it is desirable to analyze the recording of heart sounds in a computerized way. The previous studies of segmentation have been reported in literature. With the reference ECG, the performance of heart sound segmentation is quite good. However, ECG is another signal source. It is maybe discommodious for medical check-up. The electronic stethoscope may become too complex if ECG circuit is embedded in. Therefore, the segmentation algorithms using heart sounds as sole source are welcome, such as [1-5]. H. Liang et al. [1] identified S1 and S2 by picking up the peaks of heart sounds envelope. Unfortunately, the peaks of envelope are often affected by noise. On the other hand, S1 and S2 may be missed if the amplitude is too small. J Vepa et al. [2] proposed a segmentation method which used energy-based and simplicity-based features computed from multi-level wavelet decomposition coefficients to identify S1 and S2. The segmentation method proposed by S Ari et al. [3] needed only the average heart rate or a rectangular window with proper width as auxiliary information, but the way to obtain the average heart rate was not mentioned in detail. An automatic segmentation is desirable. The objective of this paper is to develop an automatic, robust algorithm for segmenting heart sound. The features of heart sounds in time domain, frequency domain and medical domain are fully exploited. The sequence of heart sound was divided cycle by cycle first. The heart sound signal of a cardiac cycle was then decomposed into a number of atoms characterized by timing delay, frequency, amplitude, time width and phase. S1 and S2 were identified by the hybrid analysis of atom clusters and medical knowledge. Cycle frequency spectrum and short time cycle frequency spectrum The cardiac cycle is a sequence of events that is repeated cyclically as a heart beats. A heart sound signal is thus quasi-cyclostationary. The fundamental concept underlying the cardiac cycle detection is to exploit the repetitive behaviors. We may estimate the fundamental cycle frequency using cyclic statistics. 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering (ICMMCCE 2015) © 2015. The authors Published by Atlantis Press 361 Cycle Frequency Spectrum A second-order cyclic statistics, proposed by Gardner, to assess cyclostationarity of a cyclic signals, ( ) x t , at cycle frequency α and frequency f is cyclic spectral autocoherence function, ( , ) x f ρ α , which is defined as 1/2 ( , ) ( , ) [ (0, / 2) (0, / 2)] x x x x S f f S f S f α ρ α α α = + − (1) ( , ) x S f α is written as 2 ( , ) ( , ) j f x x S f R e d π τ α α τ τ +∞ −

Automatic Heart Sound Segmentation Method Based on Cyclostationarity and Clustering