We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Adventitious and Normal Respiratory Sound Analysis with Machine Learning Methods.
- Authors
Demirci, Burcu Acar; Koçyiğit, Yücel; Kızılırmak, Deniz; Havlucu, Yavuz
- Abstract
Computerized respiratory sound analysis systems provide vital information regarding the current condition of the lung. These systems, used by physicians for the diagnosis of various respiratory diseases, help to classify respiratory sounds. Since physicians have differing degrees of knowledge and experience, this can cause differences in diagnosis and therefore treatment. Well-calibrated machine learning tools can help physicians make more informed decisions. For this purpose, different machine learning classifiers and feature extraction models have been developed to classify respiratory sounds from healthy individuals and patients. In this study, the combinations of Empirical Mode Decomposition, Mel Frequency Cepstral Coefficients, and Wavelet Transform methods are used for feature extraction, and k-Nearest Neighbor, Artificial Neural Networks, and Support Vector Machines are used for classification. The highest accuracy has been achieved as 98.8% when Mel Frequency Cepstral Coefficient and k-Nearest Neighbor methods are used in combination.
- Subjects
RESPIRATORY diseases; MACHINE learning; FEATURE extraction; WAVELET transforms; SCANNING electron microscopes
- Publication
Celal Bayar University Journal of Science, 2022, Vol 18, Issue 2, p169
- ISSN
1305-310X
- Publication type
Article
- DOI
10.18466/cbayarfbe.1002917