Classification of Healthy and Pathologic Lung Sounds Recorded with Electronic Auscultation

Aras S., GANGAL A., BÜLBÜL Y.

23nd Signal Processing and Communications Applications Conference (SIU), Malatya, Turkey, 16 - 19 May 2015, pp.252-255 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/siu.2015.7129807
  • City: Malatya
  • Country: Turkey
  • Page Numbers: pp.252-255
  • Keywords: Electronic Auscultation, Lung Sounds, Mel Frequency Cepstrum Coefficients, Linear Predictive Coding, RESPIRATORY SOUNDS, INSTRUMENT, DSP
  • Karadeniz Technical University Affiliated: Yes

Abstract

In this paper, lung sounds recorded by electronic auscultation, were classified as healthy and pathological. Linear Predictive Coding coefficients, Mel Frequency Cepstrum Coefficients mean and standard deviation were used as features. Records are consists of two data sets containing different respiratory cycle, k-Nearest Neighbor classification algorithm used and the performance obtained were discussed in the conclusion to the case of using different data sets and different attributes.