Aplicación de técnicas de clustering en sonidos adventicios para mejorar la interpretabilidad y detección de estertores

Application of clustering techniques for lung sounds to improve interpretability and detection of crackles


Thumbnail
View/Open
Date
2015-01-05
Author
Sosa, Germán D
Velásquez Clavijo, Fabián
Metadata
Show full item record
Compartir:

Citar con el siguiente link: http://hdl.handle.net/11323/1564
https://revistascientificas.cuc.edu.co/ingecuc/article/view/366

Abstract

Due to the subjectivity involved currently in pulmonary auscultation process and its diagnostic to evaluate the condition of respiratory airways, this work pretends to evaluate the performance of clustering algorithms such as k-means and DBSCAN to perform a computational analysis of lung sounds aiming to visualize a representation of such sounds that highlights the presence of crackles and the energy associated with them. In order to achieve that goal, Wavelet analysis techniques were used in contrast to traditional frequency analysis given the similarity between the typical waveform for a crackle and the wavelet sym4. Once the lung sound signal with isolated crackles is obtained, the clustering process groups crackles in regions of high density and provides visualization that might be useful for the diagnostic made by an expert. Evaluation suggests that k-means groups crackle more effective than DBSCAN in terms of generated clusters.
 
Debido a la subjetividad que involucra actualmente el proceso de auscultación pulmonar y su diagnóstico para evaluar la condición de las vías respiratorias de un paciente, este trabajo busca evaluar el desempeño de los algoritmos de clustering: k-means y DBSCAN para efectuar un análisis computacional de sonidos pulmonares con el objetivo de visualizar una representación de dichos sonidos que exalte la presencia de estertores y la energía contenida en ellos. Para este fin, se emplearon técnicas de descomposición y análisis Wavelet a diferencia del tradicional análisis en frecuencia dada la similitud entre la forma de onda de un estertor típico y la wavelet sym4. Obtenida la señal de sonido pulmonar con estertores aislados, el proceso de clustering agrupa estertores en regiones de alta presencia y ofrece una visualización que puede ser de utilidad para el diagnóstico hecho por un experto. La evaluación hecha sugiere que k-means agrupa conjuntos de estertores de forma más efectiva que DBSCAN en términos de clusters generados.
 
Referencias:
[1] W. H. Organization et al., “The top 10 causes of death: The 10 leading causes of death in the world, 2000 and 2011,” 2013. [2] M. Munakata, H. Ukita, I. Doi, Y. Ohtsuka, Y. Masaki, Y. Homma, and Y. Kawakami, “Spectral and waveform characteristics of fine and coarse crackles.” Thorax, vol. 46, no. 9, pp. 651–657, 1991. DOI:10.1136/thx.46.9.651 [3] M. Yeginer and Y. P. Kahya, “Elimination of vesicular sounds from pulmonary crackle waveforms,” Computer methods and programs in biomedicine, vol. 89, no. 1, pp. 1–13, 2008. DOI:10.1016/j.cmpb.2007.10.002 [4] T. Kaisia, A. Sovijarvi, P. Piirila, H. Rajala, S. Haltsonen, and T. Rosqvist, “Validated method for automatic detection of lung sound crackles,” Medical and Biological Engineering and Computing, vol. 29, no. 5, pp. 517–521, 1991. DOI:10.1007/BF02442324 [5] M. Yeginer and Y. P. Kahya, “Probing the existence of medium pulmonary crackles via model-based clustering,” Computers in biology and medicine, vol. 40, no. 9, pp. 765–774, 2010. DOI:10.1016/j.compbiomed.2010.07.004 [6] M. Paoletti, G. Camiciottoli, E. Meoni, F. Bigazzi, L. Cestelli, M. Pistolesi, and C. Marchesi, “Explorative data analysis techniques and unsupervised clustering methods to support clinical assessment of chronic obstructive pulmonary disease (COPD) phenotypes,” Journal of biomedical informatics, vol. 42, no. 6, pp. 1013–1021, 2009. DOI:10.1016/j.jbi.2009.05.008 [7] X. Lu and M. Bahoura, “An automatic system for crackles detection and classification,” in Electrical and Computer Engineering, 2006. CCECE’06. Canadian Conference. IEEE, 2006, pp. 725–729. DOI:10.1109/CCECE.2006.277698 [8] M. Bahoura and X. Lu, “Separation of crackles from vesicular sounds using wavelet packet transform,” in Acoustics, Speech and Signal Processing, 2006. ICASSP 2006 Proceedings. 2006 IEEE International Conference, vol. 2. IEEE, 2006, pp. II–II. DOI:10.1109/ICASSP.2006.1660533 [9] L. J. Hadjileontiadis and S. M. Panas, “Separation of discontinuous adventitious sounds from vesicular sounds using a wavelet-based filter,” Biomedical Engineering, IEEE Transactions on, vol. 44, no. 12, pp. 1269–1281, 1997. DOI:10.1109/10.649999 [10] S. Mallat, A wavelet tour of signal processing. Academic Press, 1999. [11] F. Pedregosa et al., “Scikit-learn: Machine learning in Python,” J Mach Learn Res, vol. 12, pp. 2825–2830, 2011. [12] E. Alpaydin, Introduction to machine learning. MIT press, 2004. [13] J. J. Ward, “Rale lung sounds 3.1 professional edition,” Respiratory Care, vol. 50, no. 10, pp. 1385–1388, 2005. [14] D. Mazzoni, M. Brubeck, and J. Haberman, “Audacity: Free audio editor and recorder”. [En línea] Disponible en: http://audacity.sourceforge.net, 2005.
URI: http://hdl.handle.net/11323/1564
https://revistascientificas.cuc.edu.co/ingecuc/article/view/366
Collections