6533b834fe1ef96bd129cafe

RESEARCH PRODUCT

Automatic screening of cardiac disorders using wavelet analysis of heart sound

Mohan MishraAnushikha SinghAlfredo Rosado MuñozMalay Kishore Dutta

subject

Sound (medical instrument)Audio signalmedicine.diagnostic_testStethoscopebusiness.industrySpeech recognitionFeature extractionPhysical examinationAuscultationlaw.inventionWaveletlawHeart soundsmedicinebusiness

description

Body auscultation is a dominant method for physical examination of human heart using conventional stethoscope. This clinical method is non invasive and efficient but it requires a medical expert to interpret the heart sound for assessment of cardiac disorders. This paper presents analysis of heart sounds in wavelet domain for automated screening of cardiac disorders. Heart sound signal is transformed in wavelet domain to find out discrimination between heart sounds recorded from healthy and anomalous patients. Discriminatory features extracted from wavelet coefficients of heart sound are subjected to machine learning for screening of cardiac disorders automatically. The proposed method for screening of cardiac disorders is tested on a comprehensive database of heart sound and results are encouraging. The experimental results indicate that proposed method achieved good accuracy for diagnosis of cardiac diseases.

yearjournalcountryeditionlanguage
2017-10-012017 4th IEEE Uttar Pradesh Section International Conference on Electrical, Computer and Electronics (UPCON)
https://doi.org/10.1109/upcon.2017.8251094