Detection of Wheezes Using a Wearable Distributed Array of Microphones

Abstract

This paper presents a wheeze detection method that uses a distributed array of microphones and can be implemented as part of a wearable health monitoring system. In order to reduce the power consumption for the wearable system, the method has been developed to operate at a sampling rate of 1000Hz, instead of 8000Hz. In the design, we use two regular air conductive microphones and a bone conductive microphone to increase the accuracy of detection and make it robust against environmental noise. The two air-conductive microphones capture breathing sound while bone-conductive microphone is placed over the manubrium of the sternum in patients to record chest wall lung sound. The simulations are conducted using lung sounds from patients with wheezes and human subjects with no wheezes under different SNR conditions. The results show that the proposed method is robust against environmental noise and has good performance on wheeze detection. The approach has been implemented onto a PDA and tested with some real data.