International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 66 - Number 17 |
Year of Publication: 2013 |
Authors: A. E. El-alfi, A. F. Elgamal, R. M. Ghoniem |
10.5120/11176-6331 |
A. E. El-alfi, A. F. Elgamal, R. M. Ghoniem . A Computer-based Sound Recognition System for the Diagnosis of Pulmonary Disorders. International Journal of Computer Applications. 66, 17 ( March 2013), 22-30. DOI=10.5120/11176-6331
This paper presents a computer-based sound recognition system for diagnosis of pulmonary disorders based on the interpretation of the lung sound signals (LSS). We propose a novel method of analysis of LSS using the Mel-frequency cepstral coefficients, the spectral and temporal parameters estimated from the frequency subbands of the discrete wavelet transform. A Linde Buzo Gray (LBG) clustering neural network model is developed for classifying the LSS to one of the six categories: normal, wheeze, crackle, squawk, stridor, or rhonchus. Experimental results demonstrate the effectiveness of the proposed system in detecting pulmonary disorders.