The main objective is the implementation of a tool that reduces the noise present in the pulmonary signal captured by a stethoscope by applying digital processing techniques. This signal may be affected by ambient noise in the room where the patient is being monitored, as well as by noise from outside (traffic, ambulances, etc.).
To achieve the stated objective, multiband spectral subtraction techniques will be applied. The different incidence that the spectral bands have in the perception of the possible symptoms of disease that the patient may present will be taken into account.
Matlab will be used for the development of this tool.
The file attached is a papper about the technique to use in the implementation.
I also have a library of sounds to work with.