Journal of Northeastern University Natural Science ›› 2017, Vol. 38 ›› Issue (11): 1534-1537.DOI: 10.12068/j.issn.1005-3026.2017.11.004

• Information & Control • Previous Articles     Next Articles

Detection of Wheeze Based on Hough Transform of Spectrogram

ZHANG Ke-xin1,2, LONG Zhe 1, WANG Xue-feng3, ZHAO Hong1   

  1. 1.School of Sino-Dutch Biomedical and Information Engineering, Northeastern University, Shenyang 110169, China; 2. The Liaoning University of Traditional Chinese Medicine, Shenyang 110847, China; 3. The First Affiliated Hospital of Liaoning University of Traditional Chinese Medicine, Shenyang 110032, China.
  • Received:2016-06-03 Revised:2016-06-03 Online:2017-11-15 Published:2017-11-13
  • Contact: ZHANG Ke-xin
  • About author:-
  • Supported by:
    -

Abstract: A method based on Hough transform to detect the wheezes from the STFT spectrum of the lung sounds was presented. This method first clips the ROI region of the STFT spectrum of the digital lung sound data, then uses the Canny operator for image edge detection, and finally identifies the wheezes automatically based on the analysis of Hough transform data. The data for clinical analysis includes the clinical acquisition of lung sounds and the international shared lung sound files. Hough transform detection method got the accuracy rate of 87% in the detection of 60 cases of wheezes, and 74% for the 70 cases of normal respiratory sound.

Key words: wheeze, Hough transform, STFT transform, spectrogram, lung sound recognition

CLC Number: