Heart Sounds Classification Using Loudness Features and Gaussian Mixture Model
Current Practice in Medical Science Vol. 5,
14 July 2022
,
Page 208-221
https://doi.org/10.9734/bpi/cpms/v5/15862D
Abstract
This paper represents a new automatic method of classifying the heart sound status using the loudness features of the heart sound. The method includes the following three main steps. First, the heart sound, which is usually found noisy, is heavily filtered by a 8th order Chebyshev Type-II filter. The event synchronous method is later used to segment the filtered heart sound into the first heart sound, systole, second heart sound and diastole. In the second step, the loudness feature is represented using the mean rows of its spectrogram. The third step categorises the heart sound using the Gaussian Mixture Model approach. With a success rate of 97.77 percent, the suggested method has been evaluated on a huge database of heart sounds containing over 3000 recordings.
- Phonocardiography
- event synchronous segmentation
- loudness
- spectrogram
- GMM classifier