Signature Extraction Technologies from Acoustic Noise of the Breathing Process in Lung Pathologies

Abstract

A simulation model of the acoustic noise of the breathing process is proposed. The statistical characteristics (spectrograms, averaged spectra and temporal realizations) of the acoustic noise of the lungs, including for some types of pathologies, are analyzed. It is proposed to distinguish the signatures of acoustic noise, considering separately different phases of the breathing process. It was shown that to determine the type of pathology, one can use a comparison of both current and averaged spectra of different breathing phases (inhalation and exhalation) of the left and right lungs and their moment characteristics.