A computer-aided MFCC-based HMM system for automatic auscultation

Authors:
Sunita Chauhan;Ping Wang;Chu Sing Lim;V. Anantharaman
Affiliations:
School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 - Nanyang Avenue, Singapore 639798, Singapore;School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 - Nanyang Avenue, Singapore 639798, Singapore;School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 - Nanyang Avenue, Singapore 639798, Singapore;Department of Emergency, Singapore General Hospital, Singapore
Venue:
Computers in Biology and Medicine
Year:
2008

Citing 2
Cited 5

A tutorial on hidden Markov models and selected applications in speech recognition

Readings in speech recognition
Training Hidden Markov Models with Multiple Observations-A Combinatorial Method

IEEE Transactions on Pattern Analysis and Machine Intelligence

Support Vectors Machine-based identification of heart valve diseases using heart sounds

Computer Methods and Programs in Biomedicine
A call-independent and automatic acoustic system for the individual recognition of animals: A novel model using four passerines

Pattern Recognition
Heart murmurs identification using random forests in assistive environments

Proceedings of the 3rd International Conference on PErvasive Technologies Related to Assistive Environments
Automatic phonocardiograph signal analysis for detecting heart valve disorders

Expert Systems with Applications: An International Journal
MCMC Bayesian inference for heart sounds screening in assistive environments

Proceedings of the 4th International Conference on PErvasive Technologies Related to Assistive Environments

Quantified Score

Hi-index	0.00

Visualization

Abstract

Auscultation, the act of listening to the sounds of internal organs, is a valuable medical diagnostic tool. Auscultation methods provide the information about a vast variety of internal body sounds originated by various organs such as heart, lungs, bowel, vascular disorders, etc. In this study, a cardiac sound registration system has been designed incorporating functions such as heart signals segmentation, classification and characterization for automated identification and ease of interpretation by the users. Considering a synergy with the domain of speech analysis, the authors introduced Mel-frequency cepstral coefficient (MFCC) to extract representative features and develop hidden Markov model (HMM) for signal classification. This system was applied to 1381 data sets of real and simulated, normal and abnormal domains. Classification rates for normal and abnormal heart sounds were found to be 95.7% for continuous murmurs, 96.25% for systolic murmurs and 90% for diastolic murmurs by a probabilistic comparison approach. This implies a high potential for the system as a diagnostic aid for primary health-care sectors.