Localizing Lung Sounds: Eigen Basis Decomposition for Localizing Sources Within a Circular Array of Sensors

摘要

Lung disorders or injury can result in changes in the production of lung sounds both spectrally and regionally. Localizing these lung sounds can provide information to the extent and location of the disorder. Difference in arrival times at a set of sensors and trian-gulation were previously proposed for acoustic imaging of the chest. We propose two algorithms for acoustic imaging using a set of eigen basis functions of the Helmholtz wave equation. These algorithms remove the sensor location contribution from the muiti sensor recordings using either an orthogonality property or a least squares based estimation after which a spatial minimum variance (MV) spectrum is applied to estimate the source locations. The use of these eigen basis functions allows possible extension to a lung sound model consisting of layered cylindrical media. Theoretical analysis of the relationship of resolution to frequency and noise power was derived and simulations verified the results obtained. Further, a Nyquist's criteria for localizing sources within a circular array shows that the radius of region where sources can be localized is inversely proportional to the frequency of sound.The resolution analysis and modified Nyquist criteria can be used for determining the number of sensors required at a given noise level, for a required resolution, frequency range, and radius of region for which sources need to be localized.