Electronic stethoscope and diagnostic algorithm

An electronic stethoscope comprising:

• a chest piece comprising:
• a hollow bell portion;
• a membrane portion covering an end of the bell portion;
• a stem portion attached to an outer circumferential wall of the hollow bell portion;
• a shoulder portion attached to the stem portion;
• and
• a microphone compartment attached to the shoulder portion;
• a microphone disposed in the microphone compartment;
• a visual indicator disposed on an exterior of the chest piece;
• and
• an electronic data connection from the microphone,
• wherein at least a portion of each of the hollow bell portion, the stem portion, the shoulder portion, and the microphone compartment form a single, continuous component composed of a polymer material and formed via a single manufacturing process.

A method of training a machine learning algorithm to identify heart and/or lung sounds comprising:

• obtaining heart and/or sound recordings corresponding to known heart and/or lung conditions and corresponding to normal healthy heart and/or lung function;
• obtaining non-heart-and/or-lung sound recordings of sounds;
• providing the heart and/or lung sound recordings and non-heart-and/or-lung sound recordings to an audio classifier model;
• segmenting the recordings via at least one of frequency, amplitude, and duration;
• inputting the segmented recordings into a U-Net neural network;
• preprocessing the segmented recordings, via the U-Net neural network, into a spectrogram for frequency equivariance, thereby producing corrected segmentation recordings;
• outputting the corrected segmentation recordings;
• and
• using the audio classifier model to train the machine learning algorithm to distinguish between the heart and/or lung sound recordings and non-heart-and/or-lung sound recordings,
• wherein the heart and/or lung sound recordings are obtained from recordings performed by an electronic stethoscope and/or a microphone, and
• wherein the audio classifier model comprises a pre-trained audio neural network (PANN), a dataset of known audio recordings, information regarding health conditions and demographics of patients with known heart and/or lung conditions, and information regarding healthy patients.