EMD based Hilbert energy spectrum analysis of adventitious lung wave signals

Abstract

The respiratory adventitious waves are analyzed effectively by time frequency analysis. In this paper, we present a new approach for rectifying the abnormality in adventitious wave. Basically, there are two types of respiratory sound waves and these are classified as wheezes and crackles. The proposed method utilizes the time frequency analysis using spectrum analysis method. The modified Empirical Mode Decomposition (EMD) called Ensemble Empirical Mode Decomposition (EEMD) to plot energy spectrum of adventitious wave is used in this work. The proposed method decomposes the respiratory adventitious wave into a different Intrinsic Mode Function (IMF). The long and short duration adventitious waves are present in a wheezing subject and this leads to production of non stationary and nonlinear sound waves. The empirical mode decomposition (EMD) decomposes such characteristic waves. The available spectrogram analyzes techniques related to additive expansions and explore amplitude wise time-frequency distribution. The methodology discussed in this context responding greatly even for correlative noise and explores energy spectra in addition to amplitude spectra. The various IMFs such produced are exhibits the fine details of adventitious wave and thus pattern can be predicted for final residual. The energy spectrum can be viewed as a diagnostic tool for accurate analysis of wheezing pattern. The decomposed frequency patterns indicate the physiological aspects. The instantaneous frequency and Hilbert energy spectrum based on above mentioned a method are employed by IMF to analyze and present the result in time-frequency distribution to explore the characteristics of inherent properties adventitious signals. The Hilbert marginal spectrum has been used to indicate overall energy distribution from each frequency component. Finally, the resultant EMD analysis along with EEMD energy spectrum is better for asthmatic subject and solves mode mixing problems.