Heart Sound Cancellation Based on Multiscale Products and Linear Prediction

Abstract

This paper presents a novel method for Heart Sound (HS) cancellation from Lung Sound (LS) records. The method uses the multiscale product of the wavelet coefficients of the original signal to detect HS-included segments. Once the HS segments are identified, the method removes them from the wavelet coefficients at every level and estimates the created gaps by using a set of linear prediction filters. It is shown that if the segment to be predicted is stationary, a final record with no audible artifacts such as clicks can be reconstructed using this approach. The results were promising for HS removal from LS records and showed no hampering of the main components of the LS. The results were confirmed both qualitatively by listening to the reconstructed signal and quantitatively by spectral analysis.

Separating Heart Sound from Lung Sound Using LabVIEW

“…Signal and noise have totally different behaviors in the wavelet domain. In [18], this behavior was analyzed using the concept of Lipschitz regularity. As an example, the Lipschitz regularity of a step function is 0.…”

“…Respiratory flow at the mouth was also recorded by a mouthpiece tube attached to a pneumotachograph and pressure transducer. Lung sound and flow signals were digitized simultaneously at 10240 Hz and 12-bits per sample (National Instruments DAQ) with custom written software in LabVIEW, while the flow signal was later decimated to 320 Hz [18].…”

“…On average it was 100 ms. In [18], estimating the removed data is done by linear prediction, using either Autoregressive (AR) or Moving Average (MA) models.…”

“…It is too simple to assume that lung sounds are stationary during the entire duration of a respiratory cycle (inspiration/expiration) especially at the vicinity of the onsets of breath. Therefore, correct selection of the order as well as the type of linear prediction model (AR or MA) must be done carefully to ensure the data used for prediction of the gaps are indeed stationary [18]. In this paper, regardless of the flow rate, ARMA modeling and prediction is used.…”

“…Signal and noise have totally different behavior in the wavelet domain. In [18], this behavior was analyzed using the concept of Lipschitz regularity. The multiplication of the DWT coefficients between the decomposition levels can lead to identification of singularities [18].…”