Abstract :
In this paper, a novel denoising tool that enhances explosive lung sounds (ELS), such as crackles and squawks (SQs), so they can more accurately and reliably be interpreted by the physicians, has been presented. This approach combines empirical mode decomposition (EMD) with fractal dimension (FD) analysis, highlighting, simultaneously, the advantageous complementarities drawn from such bridging, through its application to experimental data. The results obtained show that this hybrid denoising scheme can achieve good discrimination between useful ELS and background noise. Finally, its adaptive performance and its low computational complexity make it applicable to other separation problems involving nonstationary transient signals contaminated by uncorrelated additive noise
Keywords :
acoustic signal processing; adaptive signal processing; bioacoustics; filtering theory; fractals; lung; medical signal processing; signal denoising; adaptive performance; background noise; computational complexity; crackle sound; empirical mode decomposition; explosive lung sounds; fractal dimension filter; hybrid denoising scheme; nonstationary transient signals; squawk sound; uncorrelated additive noise; Acoustic signal detection; Background noise; Carbon capture and storage; Explosives; Filters; Fractals; Frequency; Lungs; Noise reduction; Noninvasive treatment;
