Title :
Classification of normalized healty and pathological lung sounds
Author :
Aras, Selim ; Gangal, Ali ; Bulbul, Yilmaz
Author_Institution :
Elektrik-Elektron. Muhendisligi Bolumu, Karadeniz Teknik Univ., Trabzon, Turkey
Abstract :
In this paper, lung sounds were classified as healthy and pathological. Means and standard deviations of Mel Frequency Cepstrum Coefficients were used as features and also parameters obtained by exponential curve fitted to these values. Optimum performance obtained as 92.65% when exponential curve fit parameters used for the average of Mel Frequency Cepstrum Coefficients. k-Nearest Neighbor algorithm were used in the classification process.
Keywords :
audio signal processing; cepstral analysis; lung; medical signal processing; signal classification; exponential curve fitted; k-nearest neighbor algorithm; mel frequency cepstrum coefficient; normalized healthy lung sound; pathological lung sound; sound classification; standard deviation; Cepstrum; Computers; Conferences; Hidden Markov models; IEEE Engineering in Medicine and Biology Society; Lungs; Medical diagnostic imaging; electronic auscultation; exponential curve fitting; lung sounds; mel frequency cepstrum coefficients;
Conference_Titel :
Signal Processing and Communications Applications Conference (SIU), 2015 23th
Conference_Location :
Malatya
DOI :
10.1109/SIU.2015.7130039
