Name
Abstract | ||
|---|---|---|
Adventitious lung sounds (ALS) as crackles and wheezes are present in different lung alterations and their automated characterization and recognition have become relevant. In fact, recently their 2D spatial distribution (SD) imaging has been proposed to help diagnose of pulmonary diseases. In this work, independent component analysis (ICA) by infomax was used to find crackles sources and from them to apply a time variant autoregressive model (TVAR) to count and imaging the ALS. The proposed methodology was assessed on multichannel LS recordings by embedding simulated fine crackles with known SD in recorded normal breathing sounds. Afterwards, the adventitious image of two patients with fibrosis and emphysema were obtained and contrasted with the classical pulmonary auscultation provided by a pneumologist. The results showed that combining ICA and TVAR leads to a robust methodology to imaging ALS. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1109/EMBC.2013.6609760 | EMBC |
Keywords | Field | DocType |
ica-tvar scheme,time variant autoregressive model,medical disorders,diseases,wheezes,robust methodology,automated characterization,multichannel ls recordings,pneumodynamics,lung,independent component analysis,lung alterations,pulmonary diseases,physiological models,autoregressive processes,2d spatial distribution imaging,crackle sources,classical pulmonary auscultation,adventitious lung sound imaging,medical image processing,normal breathing sounds,distribution functions,imaging,graphical models | Autoregressive model,Crackles,Lung,Computer science,Speech recognition,Independent component analysis,Breathing,Auscultation,Infomax | Conference |
Volume | ISSN | Citations |
2013 | 1557-170X | 1 |
PageRank | References | Authors |
0.35 | 4 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| S Castañeda-Villa | 1 | 1 | 0.35 |
| N Castaneda-Villa | 2 | 1 | 0.35 |
| Ramón González-Camarena | 3 | 46 | 8.33 |
| M Mejia-Avila | 4 | 2 | 1.40 |
| T Aljama-Corrales | 5 | 40 | 7.43 |