TY - JOUR
T1 - Automatic phonocardiograph signal analysis for detecting heart valve disorders
AU - Kao, Wen Chung
AU - Wei, Chih Chao
N1 - Funding Information:
This work was supported in part by the National Science Council, R.O.C., under Grants NSC 96-2221-E-003-013-MY3. The authors would like to thank Prof. Chih-Jen Lin along with his research team members Tzu-Kuo Huang and Rong-En Fan at National Taiwan University, Taiwan, R.O.C. for kindly providing the LIBSVM tool and offering valuable discussions. The authors would also like to thank Ms. Yu-Ning Chen for helping to revise the paper at National Taiwan Normal University.
PY - 2011/6
Y1 - 2011/6
N2 - Skilled cardiologists probe heart sounds by electronic stethoscope through human ears, but interpretations of heart sounds is a very special skill which is quite difficult to teach in a structured way. Because of this reason, automatic heart sound analysis in computer systems would be very helpful for medical staffs. This paper presents a complete heart sound analysis system covering from the segmentation of beat cycles to the final determination of heart conditions. The process of heart beat cycle segmentation includes autocorrelation for predicting the cycle time of a heart beat. The feature extraction pipeline includes stages of the short-time Fourier transform, the discrete cosine transform, and the adaptive feature selection. Many features are extracted, but only a few specific ones are selected for the classification of each hyperplane based on a systematic approach. The experiments are done by a public heart sound database released by Texas Heart Institute. A very promising recognition rate has been achieved.
AB - Skilled cardiologists probe heart sounds by electronic stethoscope through human ears, but interpretations of heart sounds is a very special skill which is quite difficult to teach in a structured way. Because of this reason, automatic heart sound analysis in computer systems would be very helpful for medical staffs. This paper presents a complete heart sound analysis system covering from the segmentation of beat cycles to the final determination of heart conditions. The process of heart beat cycle segmentation includes autocorrelation for predicting the cycle time of a heart beat. The feature extraction pipeline includes stages of the short-time Fourier transform, the discrete cosine transform, and the adaptive feature selection. Many features are extracted, but only a few specific ones are selected for the classification of each hyperplane based on a systematic approach. The experiments are done by a public heart sound database released by Texas Heart Institute. A very promising recognition rate has been achieved.
KW - 2-D discrete cosine transform
KW - Adaptive feature extraction
KW - Phonocardiogram (PCG)
KW - Short-time Fourier transform
KW - Support vector machines (SVMs)
UR - http://www.scopus.com/inward/record.url?scp=79951578117&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=79951578117&partnerID=8YFLogxK
U2 - 10.1016/j.eswa.2010.11.100
DO - 10.1016/j.eswa.2010.11.100
M3 - Article
AN - SCOPUS:79951578117
SN - 0957-4174
VL - 38
SP - 6458
EP - 6468
JO - Expert Systems with Applications
JF - Expert Systems with Applications
IS - 6
ER -