Lung sound denoising stethoscope, algorithm, and related methods

A method of processing signals detected by an electronic stethoscope, the method comprising:

• obtaining an auscultation signal from a body under observation with said electronic stethoscope, said auscultation signal comprising a target body sound;
• obtaining a noise signal comprising noise from an environment of said body, wherein said auscultation signal and said noise signal are detected substantially simultaneously and in-phase;
• and
• obtaining a resultant signal by subtracting adjusted noise from said auscultation signal, said resultant signal comprising said target body sound,
• wherein said adjusted noise signal is calculated based on at least a portion of said noise signal adjusted by a plurality of spectral weighting factors across a plurality of time windows and a plurality of frequency bands, wherein each of said spectral weighting factors is individually selected at each frequency band to block predetermined higher frequency ranges of said noise signal compared to a lower frequency range corresponding to said target body sound,
• wherein the plurality of spectral weighting factors comprise a signal-to-noise ratio (SNR) weighting factor for each of the plurality of frequency bands and a spectral weighting factor in a frequency domain across the plurality of frequency bands selected to enhance the auscultation signal at one or more first frequency bands in the plurality of frequency bands and penalize the auscultation signal at one or more second frequency bands different from the one or more first frequency bands in the plurality of frequency bands depending on a spectral profile of the noise signal.

A non-transitory computer-readable medium comprising software, which when executed by a computer causes the computer to:

• receive a first signal from an electronic stethoscope monitoring a body, said first signal comprising a target body sound;
• receive a second signal comprising noise, wherein said first signal and said second signal are detected substantially simultaneously and in-phase;
• and
• obtain a resultant signal by subtracting adjusted second signal from said first signal, said resultant signal comprising the target body sound,
• wherein said adjusted second signal is calculated based on at least a portion of said second signal adjusted by a plurality of spectral weighting factors across a plurality of time windows and a plurality of frequency bands, wherein each of said spectral weighting factors is individually selected at each frequency band to block predetermined higher frequency ranges of said second signal compared to a lower frequency range corresponding to said target body sound,
• wherein the plurality of spectral weighting factors comprise a signal-to-noise ratio (SNR) weighting factor for each of the plurality of frequency bands and a spectral weighting factor in a frequency domain across the plurality of frequency bands selected to enhance the auscultation signal at one or more first frequency bands in the plurality of frequency bands and penalize the auscultation signal at one or more second frequency bands different from the one or more first frequency bands in the plurality of frequency bands depending on a spectral profile of the noise signal.