| Presentation | 2014-05-22 Signal Processing for Lung-Sound Recognition : Separation and Classification of Continuous and Discontinuous Adventitious Sounds Tomoya SAKAI, Takahiro UCHIDA, Aeri SAKAGUCHI, Senya KIYASU, Sueharu MIYAHARA, Mikio OKA, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | This paper presents signal separation techniques for pattern recognition of lung sounds. From a viewpoint of pattern recognition, pulmonary auscultation is essentially a multi-label classification problem where all lung-sound components are to be predicted for an auscultatory sound. An auscultatory sound acquired from a patient with pulmonary disorder may be composed of multiple types of adventitious (abnormal) sounds as well as breath. The present approach decomposes the auscultatory sound into components without explicitly modeling them. Robust principal component analysis is applied to the sound spectrogram to yield sound components with low-rank and sparse spectrograms. It is experimentally shown that continuous adventitious sound component is obtained from the sparse spectrogram. Coarse and fine crackles can be extracted and classified by sparse representation of the sound component from the low-rank spectrogram. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Respiratory system diagnosis / Electronic auscultation / Signal separation |
| Paper # | SIP2014-11,IE2014-11,PRMU2014-11,MI2014-11 |
| Date of Issue |
| Conference Information | |
| Committee | MI |
|---|---|
| Conference Date | 2014/5/15(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Medical Imaging (MI) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Signal Processing for Lung-Sound Recognition : Separation and Classification of Continuous and Discontinuous Adventitious Sounds |
| Sub Title (in English) | |
| Keyword(1) | Respiratory system diagnosis |
| Keyword(2) | Electronic auscultation |
| Keyword(3) | Signal separation |
| 1st Author's Name | Tomoya SAKAI |
| 1st Author's Affiliation | Nagasaki Universiy() |
| 2nd Author's Name | Takahiro UCHIDA |
| 2nd Author's Affiliation | Nagasaki Universiy |
| 3rd Author's Name | Aeri SAKAGUCHI |
| 3rd Author's Affiliation | Nagasaki Universiy |
| 4th Author's Name | Senya KIYASU |
| 4th Author's Affiliation | Nagasaki Universiy |
| 5th Author's Name | Sueharu MIYAHARA |
| 5th Author's Affiliation | / Kawasaki Medical School |
| 6th Author's Name | Mikio OKA |
| 6th Author's Affiliation | |
| Date | 2014-05-22 |
| Paper # | SIP2014-11,IE2014-11,PRMU2014-11,MI2014-11 |
| Volume (vol) | vol.114 |
| Number (no) | 42 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |