Design of a Robust Lung Sound Acquisition System for Reliable Acoustic Lung Imaging

Chronic lung diseases are typically tied to acoustic airway obstruction, and acoustic imaging transmuted from lung sounds from an array of digital sensors at different locations is used as an alternative for frequent lung function assessment. Digital stethoscopes are a widely used clinical tool for identifying airway obstruction through lung signals. Nevertheless, digital stethoscopes are expensive and remain riddled with several issues that limit their signal quality in noisy settings and result in unreliable acoustic lung imaging. Thus, the design and validation of a reconfigurable wearable and low-cost yet robust lung sound acquisition system with microelectromechanical systems microphones that enable continual lung function assessment and are robust to noise by reducing external noise contamination through hardware redesign and dynamic signal processing was proposed in this study. The system is objectively compared to commercially available digital devices in a simulated noisy setting and quantified using root mean square error (RMSE), reflecting the accuracy in capturing desired signals and signal-to-noise ratio (SNR), reflecting the robustness to noise contamination. The system achieved better results than commercial digital sensors in terms of RMSE by about 0.15 and SNR by about 7 dB. Our system sensor’s position is also robust in representing lung sound signals in terms of sensing sensitivity power spectrum mapping for frequencies ranging from 100 to 2000 Hz, achieving a power ratio loss of about 6 dB compared to the other digital devices of about 10 dB, at 20 mm away from the sensor’s center.