1. Pneumonia diagnosis algorithm based on room impulse responses using cough sounds
- Author
-
Jin Yong Jeon, Na Young Kim, Sang Heon Kim, Hong Jin Kim, and Gwun Il Park
- Subjects
Acoustics and Ultrasonics ,Arts and Humanities (miscellaneous) - Abstract
For data augmentation of the pneumonia diagnosis algorithm by deep learning, an image conversion method through a convolutional method utilizing spatial impulse and sound quality factors of cough sounds is proposed. First, reverberant spaces with different volumes were implemented, and spatial impulse responses were generated for each space through computer simulation of spatial models according to sound source and receiver points. Sound quality analysis was performed to improve accuracy, and 2-D sound quality data of time series was converted into 3-D image data using the Gramian Angular Field (GAF) method for combination between heterogeneous data. As a result, 97.5% accuracy was obtained for the configured dataset. The result of this study is expected to be used to develop diagnostic algorithms for various respiratory diseases including pneumonia in the future.
- Published
- 2022