Heart Sound Analysis for Diagnosis of Heart Diseases in Newborns

Many studies have been conducted in recent years to automatically differentiate normal heart sounds from heart sounds with pathological murmurs using audio signal processing in early stage. Serious cardiac pathology may exist without symptoms. The purpose of this study is developing an automatic heart sound signal analysis system, able to support the physician in the diagnosing of heart murmurs at early stage of life. Heart murmurs are the first signs of heart disease. We screened newborns for normal (innocent) and pathological murmurs. This paper presents an analysis and comparisons of spectrograms after smoothing phonocardiogram signals (PCG) with Cepstrum, Bispectrum, and Wigner Bispectrum techniques. A comparison between these methods has shown that higher order spectra, as Bispectrum and Wigner bispectrum, gave the best results.