Performance Metrics Analysis of Adaptive threshold Empirical Mode Decomposition Denoising method for suppression of noise in Lung sounds

Chest auscultation is a non-invasive and widely used tool for the detection of pulmonary disease. Though it is a powerful method of pulmonary function test, it relics challenges in different issues that bound its diagnostic ability. Mainly in motion artifacts, other environmental noise, and heart sound interference, hence contaminates the original content of the lung sound. This study proposes an adaptive threshold Empirical Mode Decomposition (aEMD) technique to denoise the lung sound signals thus improving the signal quality for the detection of respiratory pathologies. The proposed scheme is refined to offset maximum noise suppression against preserving the reliability of the lung sound signal. With this algorithm, the maximum Signal to Noise ratio (SNR) of 43.89 dB, Correlation Coefficient of 0.995, and minimum Root Mean Square Error (RMSE) of 0.00122 is achieved. Further, it can be implemented in real-time to assist medical experts doctors to make clear interpretations of the respiratory sound-related disorder.